5.1 MIMO Antenna Design

University of Bradford eThesis

This thesis is hosted in Bradford Scholars – The University of Bradford Open Access repository. Visit the repository for full metadata or to contact the repository team

INVESTIGATION OF INTEGRATED

DECOUPLING METHODS FOR MIMO ANTENNA

SYSTEMS

A. M. S. SALEH

PHD

2019 Investigation of Integrated Decoupling Methods for MIMO Antenna Systems

Design, Modelling and Implementation of MIMO Antenna Systems for Different Spectrum Applications with High Port-to-Port Isolation Using Different Decoupling Techniques

Adham Maan Saleh SALEH

Submitted for the Degree of

Doctor of Philosophy

Faculty of Engineering and Informatics

University of Bradford

2019 ABSTRACT

INVESTIGATION OF INTEGRATED DECOUPLING METHODS FOR MIMO ANTENNA SYSTEMS Design, Modelling and Implementation of MIMO Antenna Systems for Different Spectrum Applications with High Port-to-Port Isolation Using Different Decoupling Techniques

Keywords

Multiple Input Multiple Output (MIMO) Antenna, Isolation, Envelope Correlation Coefficient (ECC), Neutralization Line (NL), Parasitic Element, Metamaterial, Multi-Band Antenna, Wireless local area network (WLAN), Worldwide Interoperability for Microwave Access (WiMAX), 5th Generation Mobile technology (5G)

Multiple-Input-Multiple-Output (MIMO) antenna technology refers to an antenna with multiple radiators at both transmitter and receiver ends. It is designed to increase the data rate in wireless communication systems by achieving multiple channels occupying the same bandwidth in a multipath environment. The main drawback associated with this technology is the coupling between the radiating elements. A MIMO antenna system merely acts as an antenna array if the coupling between the radiating elements is high. For this reason, strong decoupling between the radiating elements should be achieved, in order to utilize the benefits of MIMO technology. The main objectives of this thesis are to investigate and implement several printed MIMO antenna geometries with integrated decoupling approaches for WLAN, WiMAX, and 5G applications. The characteristics of MIMO antenna performance have been reported in terms of scattering parameters, envelope correlation coefficient (ECC), total active reflection coefficient (TARC), channel capacity loss (CCL), diversity gain (DG), antenna efficiency, antenna peak gain and antenna radiation patterns. Three new 2×2 MIMO array antennas are proposed, covering dual and multiple spectrum bandwidths for WLAN (2.4/5.2/5.8 GHz) and WiMAX (3.5 GHz) applications. These designs employ a combination of DGS and neutralization line methods to reduce the coupling caused by the surface current in the ground plane and between the radiating antenna elements. The minimum achieved isolation between the MIMO antennas is found to be better than 15 dB and in some bands exceeds 30 dB. The matching impedance is improved and the correlation coefficient values achieved for all three antennas are very low. In addition, the diversity gains over all spectrum bands are very close to the ideal value (DG = 10 dB). The forth proposed MIMO antenna is a compact dual-band MIMO antenna operating at WLAN bands (2.4/5.2/5.8 GHz). The antenna structure consists of two concentric double square rings radiating elements printed symmetrically. A new method is applied which combines the defected ground structure (DGS) decoupling method with five parasitic elements to reduce the coupling between the radiating antennas in the two required bands. A metamaterial-based isolation enhancement structure is investigated in the fifth proposed MIMO antenna design. This MIMO antenna consists of two dual-band arc-shaped radiating elements working in WLAN and Sub-6 GHz 5th generation (5G) bands. The antenna placement and orientation decoupling method is applied to improve the isolation in the second band while four split-ring resonators (SRRs) are added between the radiating elements to enhance the isolation in the first band. All the designs presented in this thesis have been fabricated and measured, with the simulated and measured results agreeing well in most cases.

i Acknowledgment

Firstly, I want to praise The Almighty Allah who lightened my way during this research.

I would also like to thank my supervisors Prof Raed Abd-Alhameed, Prof Peter S.

Excell and Prof Neil J. McEwan for supporting, inspiring and guiding me to complete this research: their recommendations were very effective to correct the direction of the work to the right way.

I am also thankful to my other supervisor Dr James M. Noras and to Dr Stephen Jones for their unlimited support and guidance. I also wish to express my gratitude to my peer colleagues for their immeasurable support and guidance through this work.

I would like to thank my sponsor the Higher Committee for Education Development in

Iraq (HCED) for their moral and financial support during this research. Also, I want to thank my employer “Ninevah University” for giving me the permission to complete my

PhD study outside the country.

I want to acknolowge my wonderful wife (Saba) and my dear kids (Ibraheem, Maryam

& Ayoob) who painted a wide smile on my face during the hard moments in this work.

Thanks wife for listening to all my complaints patiently throughout my PhD study and during the writing up stage of this work. This PhD work would not be possible without your love and sacrifices.

Finally, I want to appreciate my mother for her extreme interest in my education and the unconditional support and the encouragement she has given me throughout my PhD study.

Table of Contents

ABSTRACT ...... i

Acknowledgment ...... ii

Table of Contents ...... iii

List of Figures ...... x

List of Tables ...... xxvi

List of Abbreviations ...... xxviii

List of Symbols ...... xxx

CHAPTER ONE

Introduction

1.1 Motivation ...... 2

1.2 Aim and Objectives ...... 3

1.3 Original Contributions to Scientific Knowledge ...... 4

1.4 Author’s Publications ...... 6

1.5 Thesis Layout...... 8

References ...... 11

CHAPTER TWO

Background and Literature Review of MIMO Antenna Systems

iii 2.1 The Concept of MIMO Technology ...... 12

2.2 Theoretical Background of MIMO Technology ...... 14

2.3 History of MIMO Antenna Systems ...... 17

2.4 Antennas for MIMO Systems ...... 18

2.5 Mutual Coupling ...... 20

2.6 Mutual Coupling Reduction Methods ...... 22

2.6.1 Antenna Placement and Orientation Method ...... 22

2.6.2 Neutralization Line Method ...... 32

2.6.3 Metamaterials Method ...... 38

2.6.4 Parasitic Element Method ...... 44

2.6.5 Defected Ground Structure Method ...... 51

2.7 Comparison of the Different Mutual Coupling Reduction Methods...... 58

2.8 MIMO Antennas by Band Classifications ...... 59

2.8.1 Narrow Band MIMO Antennas with Reduced Mutual Coupling ...... 59

2.8.2 Wide Band MIMO Antennas with Reduced Mutual Coupling ...... 61

2.8.3 Multi-Band MIMO Antennas with Reduced Mutual Coupling ...... 62

2.9 Hybrid Method for Mutual Coupling Reduction ...... 64

2.10 Identification of Gaps in Knowledge (Problem Statement) ...... 71

2.11 Fundamental Parameters of MIMO Antenna Systems ...... 72

2.11.1 Total Active Reflection Coefficient ...... 73

2.11.2 Isolation ...... 75 iv

2.11.3 Correlation Coefficient ...... 75

2.11.4 System Channel Capacity ...... 77

2.11.5 Mean Effective Gain ...... 80

2.11.6 Diversity Gain ...... 81

2.12 Conclusions ...... 83

References ...... 85

CHAPTER THREE

Mutual Coupling Reduction of Dual-Band Printed MIMO Antenna Using

Neutralization Line Technique and Defected Ground Structures

3.1 Single Antenna Design ...... 106

3.2 2 2 MIMO Antenna Design ...... 112

3.3 Neutralization Line ...... 114

3.3.1 S-Parameters of the Proposed MIMO Antenna with and without NL ...... 115

3.3.2 Parametric Study ...... 117

3.3.2.1 The Effect of the Position of the Neutralization Line on the S-Parameters ...... 118

3.3.2.2 The Effect of the Width of the NL on the S-parameters of the Proposed Antenna...... 120

3.4 Defected Ground Structure ...... 122

3.4.1 S-Parameters of the Proposed Antenna with DGS ...... 123

3.5 Evaluation of Antenna Efficiency and Gain ...... 125

3.6 Results and Discussions ...... 127

3.7 Comparison with Published Works...... 135

3.8 Conclusions ...... 136

References ...... 138

CHAPTER FOUR

Compact Tri-Band MIMO Antenna with High Port Isolation for WLAN and

WiMAX Applications using Neutralization Line and Defected Ground Structures

4.1 Antenna Design ...... 144

4.2 Evaluation of S-Parameters of the Proposed MIMO Antenna ...... 146

4.3 Evaluation of Antenna Gain and Efficiency ...... 149

4.4 Results and Discussion ...... 150

4.4.1 S-Parameters of MIMO Antenna With and Without NL ...... 150

4.4.2 Diversity Performance of the Proposed Antenna ...... 154

4.5 Comparison with Published Works ...... 160

4.6 Conclusions ...... 161

References ...... 163

CHAPTER FIVE

Dual-Band Printed MIMO Antenna Decoupling Based on Defected Ground

Structures and Parasitic Elements Techniques

5.1 MIMO Antenna Design ...... 166

5.2 Parametric Study ...... 167

5.2.1 S-Parameters of the Proposed MIMO Antenna with DGS ...... 168

5.2.2 S-Parameters of the Proposed MIMO Antenna with Different Slot Heights ...... 170

5.3 Parasitic Elements ...... 172

5.3.1 Parametric Study of the Proposed Antenna with Parasitic Elements ...... 173

5.3.1.1 S-Parameters of the Proposed Antenna with Variations of Lp ...... 174

5.3.1.2 S-Parameters of the Proposed Antenna with Different Variations of Wp ...... 175

5.3.1.3 S-Parameters of the Proposed Antenna with Different Variations of W ...... 175

5.3.1.4 S-Parameters of the Proposed Antenna with and without Parasitic Elements ...... 178

5.4 Evaluation of Antenna Efficiency and Gain ...... 179

5.5 The Effect of the Ground Connection ...... 181

5.6 Hardware Realization and Comparison with Simulations ...... 182

5.7 Comparison with Published Works...... 190

5.8 Conclusions ...... 191

References ...... 193

CHAPTER SIX

Isolation Enhancement of Compact Dual-Band Printed MIMO Antenna Using an

Integration of Antenna Placement and Orientation with Metamaterial Resonators

Techniques

6.1 Single Antenna Design ...... 197

6.2 Two Elements MIMO Antenna Design ...... 201

6.2.1 Antenna Placement and Orientation ...... 202

vii 6.2.2 Metamaterials ...... 205

6.2.2.1 Theoretical Aspect...... 206

6.2.2.2 Metamaterial Unit Cell Specification and Design ...... 208

6.2.2.3 Antenna Layout with Metamaterial Resonators ...... 211

6.2.2.3 Parametric Study ...... 212

6.2.2.3.1 S-Parameters of The Proposed Antenna with and without SRRs ...... 212

6.2.2.3.1 S-Parameters of The Proposed MIMO Antenna With Three Variations of

Metamaterial Resonators ...... 214

6.3 Evaluation of Antenna Efficiency and Gain ...... 215

6.3.1 Efficiency Calculations ...... 215

6.3.2 Peak Gain Calculations ...... 217

6.4 Equivalent Circuit Model ...... 218

6.4 Results and Discussions ...... 220

6.5 Comparison with Published Works ...... 227

6.6 Conclusions ...... 228

References ...... 230

CHAPTER SEVEN

Isolation Enhancement of MIMO Antenna Using Defected Ground Structures and

Neutralization Line Techniques

7.1 The Antenna Layout ...... 236

7.2 Parametric Study ...... 237 viii 7.2.1 The Impacts of the DGS Technique on the S-Parameters ...... 238

7.2.1.1 The Influence of the First Defected Area on the S-Parameters Characteristics...... 238

7.2.1.2 The Influence of the Second Defected Areas on the S-Parameters Characteristics ...... 239

7.2.1.3 The Impact of the Three Defected Areas on the S-Parameters Characteristics ...... 241

7.2.2 The Influences of the NL on the S-Parameters ...... 242

7.2.3 The Effects of the Stub on the S-Parameters of the Antenna ...... 248

7.3 Evaluation of Antenna Efficiency and Gain ...... 249

7.4 Results and Discussions ...... 251

7.4.1 S-Parameters of the Antenna ...... 252

7.4.2 Diversity Performances of the Antenna ...... 253

7.4.3 Radiation Patterns of the Antenna ...... 256

7.5 Comparison with Published Works...... 258

7.6 Conclusions ...... 260

References ...... 261

CHAPTER EIGHT

8.1 Future Work ...... 266

References ...... 268

ix List of Figures

Fig. 2.1. Configuration of different multiple antenna systems...... 14

Fig.2.2. MIMO System block diagram ...... 15

Fig. 2.3. Configurations of V-BLAST System [11]...... 18

Fig. 2.4. Three different configurations of the proposed dual array. [36]...... 24

Fig. 2.5. The geometry of the two folded antennas [37]...... 24

Fig. 2.6. The schematic structure of the four-element MIMO antenna with different element arrangements; (a) Front side, (b) bottom side [38]...... 26

Fig. 2.7. The five different cases of an IFA array [39]...... 26

Fig. 2.8. Photograph of the modified multiply-fed shielded loop antenna with an E-dipole

[40]...... 27

Fig. 2.9. The measured S-parameters of the fabricated MIMO antenna in [41]; (a) Case

1, (b) Case 2, and (c) Case 3...... 27

Fig. 2.10. The MIMO antenna proposed in [42]; (a) The geometry of the MIMO antenna

(b) The geometry of the single radiator...... 28

Fig. 2.11. Photograph of the fabricated microstrip annular slots MIMO antenna; (a) Front side, (b) Bottom side...... 29

Fig. 2.12. The prototype of the 4-element orthogonally arranged MIMO antenna; (a) Front side, (b) Bottom side [44]...... 29

Fig. 2.13. The proposed four-element UWB MIMO antenna with different arrangements;

(a) Schematic structure, (b) Fabricated antenna [45]...... 30

Fig. 2.14. The proposed MIMO antenna showing the six different configurations with their corresponding S-parameters [46]...... 30

Fig. 2.15. Configurations of the DCS and UMTS PIFAs [47]...... 33

x Fig. 2.16. The prototype of the proposed rectangular patch MIMO antenna operating at

5.75 GHz with NL [48]...... 33

Fig. 2.17. Proposed MIMO antenna with NL for wireless USB dongle application; (a)

Geometry of the proposed antenna, (b) Prototype of the proposed antenna [49]...... 33

Fig. 2.18. A prototype of the proposed MIMO antenna with NL near to the feeding point

[50]...... 34

Fig. 2.19. Configurations of the crescent-shaped MIMO antenna with zigzag NL [51]; (a)

Top view, (b) Bottom view...... 35

Fig. 2.20. Printed MIMO antenna with wideband NL [52]; (a) Top view, (b) Bottom view.

...... 35

Fig. 2.21. Design configuration of the proposed antenna with a modified 2nd order Hilbert fractal shaped NL [54]...... 36

Fig. 2.22. Schematic structure of the proposed MIMO antenna with NL and two series lumped capacitors for tablet computer [55]...... 37

Fig. 2.23. The geometry of the proposed MIMO antenna showing the three different cases of NL shapes [56]. Finally, a summary of the state-of-art of neutralization line methods, containing the important previous works is given in Table 2.2, which details the centre frequency, isolation, electrical size, separation distance between the radiating elements, antenna peak gain, antenna efficiency and ECC...... 37

Fig. 2.24. Configuration of the proposed MIMO antenna with metamaterials in the form of capacitively-loaded loops [57]...... 39

Fig. 2.25. Configurations of the proposed L-shaped slots MIMO antenna; (a) Top view,

(b) Bottom view, (c) Side view, (d) Radiating element, (e) SRR unit cell structure [58].

...... 39

xi Fig. 2.26. Configurations of the proposed two loop radiators with metamaterial unit cells;

(a) Top view, (b) Bottom view [59]...... 40

Fig. 2.27. Printed MIMO antenna with SRR; (a) Top view, (b) Bottom view [60]...... 40

Fig. 2.28. Configuration of the proposed MIMO antenna with different positions and orientations of metamaterial unit cells[61]...... 41

Fig. 2.29. A prototype of the proposed MIMO antenna with a double-layer “mushroom” isolator [62]...... 41

Fig. 2.30. Configuration of the proposed MIMO antenna with metamaterial polarization- rotator wall [63]...... 42

Fig. 2.31. The fabricated prototype of the proposed MIMO antennas showing capacitively loaded loop metamaterial [66]...... 43

Fig. 2.32. A prototype of the four-element MIMO antenna with parasitic elements [67].

...... 45

Fig. 2.33. Configuration of the proposed MIMO antennas for USB WLAN dongle [68].

...... 45

Fig. 2.34. Configuration of the proposed MIMO antenna with folded resonator above the antenna [69]...... 46

Fig. 2.35. Configuration of the proposed MIMO antenna with dual-band resonator as a parasitic element [70]...... 46

Fig. 2.36. Configuration of the proposed MIMO antenna with parasitic elements [71]. 47

Fig. 2. 37. Fabrication of the proposed MIMO antenna with a parasitic element in the form of a meandered slot cut [72]...... 47

Fig. 2.38. Fabrication of the proposed directive element MIMO antenna using four broadband printed elements with parasitic sub-elements [74]...... 49

xii Fig. 2.39. The geometry of the proposed MIMO antenna with parasitic element; (a) Top view, (b) Rear view [75]...... 49

Fig. 2.40. The fabricated design of the proposed MIMO antenna with parasitic element;

(a) Top layer, (b) Rear layer [76]...... 49

Fig. 2.41. The geometry of the proposed MIMO antenna; (a) Top layer, (b) bottom layer

[77]...... 50

Fig. 2.42. Configuration of the two PIFAs with DGS [78]...... 52

Fig. 2.43. The schematic structure of the four-element MIMO antenna with a series of slits etched in the ground plane: (a) Side view, (b) Top view, (c) Bottom view [79]. .... 53

Fig. 2.44. The configuration of the proposed MIMO antenna with DGS in the form of bent slits and two triangular cut-outs [80]...... 53

Fig. 2.45. The schematic structure of the proposed UWB MIMO antenna with DGS in the form of a T-shaped slot [81]...... 54

Fig. 2.46. The configuration of the proposed mmWave T- shaped MIMO antenna: (a)

Top view of the simulated prototype, (b) Bottom view of the simulated prototype, (c) Top view of the fabricated prototype, (d) Bottom view of the fabricated prototype [82]. 54

Fig. 2.47. Fabricated prototypes of the MIMO antenna proposals [83]; (a) 2×2 MIMO antenna and (b) 4×4 MIMO antenna...... 55

Fig. 2.48. The prototype of the proposed 2×2 dual-polarized microstrip patch MIMO antenna; (a) Top view (b) Bottom view [84]...... 55

Fig. 2.49. MIMO antenna prototype; (a) Top layer (b) Bottom layer [85]...... 56

Fig. 2.50. Configuration of the proposed MIMO antenna integrated with two slotted microstrip resonators [86]...... 56

Fig. 2.51. Fabricated prototype of the proposed MIMO antenna with cross stubs [130].

...... 66

xiii Fig. 2.52. Schematic structure of the proposed dual-element MIMO antenna with ground stubs [131]...... 66

Fig. 2.53. Fabricated prototype of the proposed compact octagonal shaped fractal MIMO antenna with ground stubs [132]...... 66

Fig. 2.54. Schematic structure of the proposed dual-band MIMO antenna with NL and

DGS [133]...... 67

Fig. 2.55. Configuration of the proposed MIMO antenna; (a) MIMO antenna, (b) PIFA element, and (c) DGS and two meander lines [134]...... 67

Fig. 2.56. Fabricated prototypes of the proposed MIMO antennas with DGS and mu- negative metamaterial; (a) Basic two port monopole antenna, (b) Circularly polarised

MIMO antenna [135]...... 68

Fig. 2.57. Schematic structure of the proposed MIMO antenna with wideband NL and

DGS [136]...... 69

Fig. 2.58. Geometry of the proposed MIMO antenna with As-EBG, T-shaped strip and

DGS [137]...... 70

Fig. 2.59. Suspended meta-surface structure consisting of periodic metamaterial cells with neutralization line decoupling structure added on the feeding lines; (a) Top view, (b)

Rear view [138]...... 70

Fig. 3.1. Geometrical configurations of the proposed single antenna...... 107

Fig. 3.2. Simulated reflection coefficients S11 of the proposed single antenna...... 108

Fig. 3.3. Variation of simulated S11 versus frequency with different length of arm1 108

Fig. 3.4. Variation of the simulated S11 versus frequency with different length of arm2.

...... 109

Fig. 3.5. Back view of the proposed single antenna showing the two defected areas. 109 xiv Fig. 3.6. Simulated reflection coefficients S11 of the proposed single antenna with different values of r (mm)...... 110

Fig. 3.7. Surface current distributions of the proposed single antenna at two different frequencies; (a) 2.4 GHz, (b) 5.5 GHz ...... 111

Fig. 3.8. Variation of calculated peak gain and efficiency versus frequency for the proposed single antenna...... 111

Fig. 3.9. The configuration of the proposed antenna; (a) Antenna with NL (b) Antenna without NL...... 113

Fig. 3.10. Surface current distributions of the proposed MIMO antenna without NL at the resonant frequency of the two required bands; (a) 2.4 GHz. (b) 5.5 GHz...... 114

Fig. 3.11. Simulated reflection coefficients S11 with and without neutralization line for the proposed MIMO antenna...... 116

Fig. 3.12. Simulated transmission coefficients S21 with and without the neutralization line for the proposed MIMO antenna...... 116

Fig. 3.13. Surface current distributions of the proposed MIMO antenna at 2.4 GHz; (a)

With NL, (b) Without NL...... 117

Fig. 3.14. Configuration of the proposed MIMO antenna showing the height (hn) and the width (wn) of the NL...... 118

Fig. 3.15. The simulated reflection coefficient of the proposed MIMO antenna with different positions for the NL...... 119

Fig. 3.16. The simulated transmission coefficient S21 of the proposed MIMO antenna with different positions of the NL...... 119

Fig. 3.17. Simulated reflection coefficient S11 of the proposed MIMO antenna with different widths of the NL...... 121

xv Fig. 3.18. Simulated transmission coefficient S21 of the proposed MIMO antenna with different widths of the NL...... 121

Fig. 3.19. Back view of the proposed MIMO antenna showing the defected areas...... 124

Fig. 3.20. Simulated reflection coefficients of the proposed MIMO antenna with different values of r (mm)...... 124

Fig. 3.21. Simulated transmission coefficients of the proposed MIMO antenna with different values of r (mm)...... 125

Fig. 3.22. Variation of calculated efficiency versus frequency for the single antenna, the

MIMO antenna without decoupling and MIMO antenna with decoupling...... 126

Fig. 3.23. Variation of calculated peak gain versus frequency for the single antenna, the

MIMO antenna without decoupling and MIMO antenna with decoupling...... 126

Fig. 3.24.The fabricated MIMO antenna design with NL (a) Top view, (b) Back view.

...... 127

Fig. 3.25. Simulated and measured S11 of the proposed MIMO antenna with decoupling methods...... 128

Fig. 3.26. Simulated and measured S21 of the proposed MIMO antenna with decoupling methods...... 128

Fig. 3.27. Simulated and measured ECC of the proposed MIMO antenna with NL. ... 129

Fig. 3.28. Simulated and measured CCL of the proposed MIMO antenna decoupling methods...... 130

Fig. 3.29. Calculated TARC with a different phase of the proposed MIMO antenna with decoupling methods...... 131

Fig. 3.30. The calculated average value of TARC for the proposed MIMO antenna with decoupling methods...... 132

xvi Fig. 3.31. Measured and simulated diversity gain of the proposed MIMO antenna decoupling methods...... 133

Fig. 3.32. Simulated and measured radiation patterns of the proposed antenna for two planes [(1 and 2) y-x plane and (3 and 4) z-x plane] at (a) 2.4 GHz (b) 5.2 GHz and (c)

5.8 GHz. Port 1 is excited and port 2 is terminated. Solid curve: simulated results, dashed curve: measured results, “1 and 3” co-polar components and “2 and 4” cross-polar components ...... 134

Fig. 3.33. The simulated 3-D pattern of the proposed MIMO antenna at (a) 2.4 GHz (b)

5.2 GHz and (c) 5.8 GHz...... 135

Fig. 4.1. Schematic structure of the proposed MIMO antenna...... 145

Fig. 4.2. Configurations of the proposed MIMO antenna showing the decoupling approaches. (a) With DGS, (b) With DGS and NL...... 146

Fig. 4.3. Simulated reflection coefficients of the proposed MIMO antenna with different values of R (mm)...... 147

Fig. 4.4. Simulated transmission coefficients of the proposed MIMO antenna with different values of R (mm)...... 147

Fig. 4.5. Simulated reflection coefficients of the proposed MIMO antenna with and without decoupling approaches ...... 148

Fig. 4.6. Simulated transmission coefficients of the proposed MIMO antenna with and without decoupling approaches ...... 148

Fig. 4.7. Variation of calculated peak gain versus frequency for the proposed MIMO antenna with and without NL...... 149

Fig. 4.8. Variation of calculated efficiency versus frequency for the proposed MIMO antenna with and without NL...... 150 xvii Fig. 4.9. The fabricated MIMO antenna design; (a) With NL, (b) Without NL...... 151

Fig. 4.10. Simulated and measured reflection coefficients S11; (a) Without NL, (b) With

NL...... 152

Fig. 4.11. Simulated and measured transmission coefficients S21 for the proposed MIMO antenna with a NL...... 153

Fig. 4.12. Surface current distributions of the proposed antenna at 3.5 GHz; (a) Without

NL; (b) With NL ...... 153

Fig. 4.13. Simulated and measured ECC of the proposed MIMO antenna with NL. ... 155

Fig. 4.14. Simulated and measured CCL of the proposed MIMO antenna with NL. ... 156

Fig. 4.15. Calculated TARC with varying phases for the proposed MIMO antenna with

NL...... 156

Fig. 4.16. The calculated average value of TARC for the proposed MIMO antenna with

NL...... 157

Fig. 4.17. The simulated diversity gain of the proposed MIMO antenna with NL...... 157

Fig. 4.18. Simulated and measured radiation patterns of the proposed antenna for two planes [(1 and 2) y-x plane and (3 and 4) z-x plane] at (a) 2.4 GHz, (b) 3.5 GHz, (c) 5.2

GHz and (d) 5.8 GHz. Port 1 is excited and port 2 is terminated. Solid curve: simulated results, dashed curve: measured results, “1 and 3” are co-polar components and “2 and

4” are cross-polar components...... 159

Fig. 4.19. The simulated 3-D pattern of the proposed MIMO antenna at (a) 2.4 GHz (b)

3.5 GHz, (c) 5.2 and (d) 5.8 GHz...... 160

Fig. 5.1. Configurations of the proposed antenna; (a) Front view; (b) Back view...... 167

Fig. 5.2. Back view of the proposed MIMO antenna showing the four defected areas.

...... 168 xviii Fig. 5.3. Simulated reflection coefficients of the proposed MIMO antenna with different values of U (mm)...... 169

Fig. 5.4. Simulated transmission coefficients of the proposed MIMO antenna with different values of U (mm)...... 169

Fig. 5.5. Back view of the proposed MIMO antenna showing the two slots...... 170

Fig. 5.6. Simulated reflection coefficients of the proposed MIMO antenna with different values of LS (mm)...... 171

Fig. 5.7. Simulated transmission coefficients of the proposed MIMO antenna with different values of LS (mm)...... 171

Fig. 5.8. Mutual coupling reduction mechanism through two parasitic elements: (a) Two radiating elements; (b) Two radiating elements with parasitic elements [4]...... 173

Fig. 5.9. Simulated reflection coefficients with different values of Lp...... 174

Fig. 5.10. Simulated transmission coefficients with different values of Lp...... 174

Fig. 5.11. Simulated reflection coefficients with different values of Wp...... 176

Fig. 5.12. Simulated transmission coefficients with different values of Wp...... 176

Fig. 5.13. Simulated reflection coefficients with different values of W...... 177

Fig. 5.14. Simulated transmission coefficients with different values of W...... 177

Fig. 5.15. Simulated reflection coefficients S11 with and without parasitic element. ... 178

Fig. 5.16. Simulated transmission coefficients S21 with and without parasitic element.

...... 179

Fig. 5.17. Variation of calculated efficiency versus frequency for MIMO antenna with and without parasitic elements...... 180

Fig. 5.18. Simulated peak gain for the MIMO antenna with and without parasitic elements...... 180

Fig. 5.19. Back view of the proposed antenna showing the two slots...... 181

xix Fig. 5.20. Simulated reflection coefficients with different values of d (cm)...... 181

Fig. 5.21. Simulated transmission coefficients with different values of d (cm)...... 182

Fig. 5.22. The fabricated MIMO antenna design with parasitic elements (a) Top view, (b)

Back view ...... 183

Fig. 5.23. Simulated and measured S11 of the proposed MIMO antenna with parasitic elements...... 183

Fig. 5.24. Simulated and measured S21 of the proposed MIMO antenna with parasitic elements...... 184

Fig. 5.25. Simulated and measured ECC of the proposed MIMO antenna with parasitic elements...... 185

Fig. 5.26. Simulated and measured CCL of the proposed MIMO antenna with parasitic elements...... 186

Fig. 5.27. Calculated TARC with different phases of the proposed MIMO antenna with parasitic elements...... 187

Fig. 5.28. The calculated average value of TARC for the proposed MIMO antenna with parasitic elements...... 187

Fig. 5.29. Simulated diversity gain of the proposed MIMO antenna...... 188

Fig. 5.30. Simulated and measured radiation patterns of the proposed antenna for two planes [(1 and 2) y-x plane and (3 and 4) z-x plane] at (a) 2.4 GHz (b) 5.2 GHz and (c)

5.8 GHz. Port 1 is excited and port 2 is terminated. Solid curves: simulated results, dashed curves: measured results, “1 and 3” are co-polar components and “2 and 4” are cross- polar components...... 189

Fig. 5.31. The simulated 3-D pattern of the proposed MIMO antenna at (a) 2.4 GHz (b)

5.2 GHz and (c) 5.8 GHz...... 190

xx Fig. 6.1. Geometrical configurations of the proposed single antenna...... 199

Fig. 6.2. Variation of simulated S11 versus frequency with different length of OPR. ... 200

Fig. 6.3. Variation of the simulated S11 versus frequency with different length of OPQ.

...... 200

Fig. 6.4. Variation of the calculated gain and efficiency for the single element antenna.

...... 201

Fig. 6.5. Four different orientations of the proposed MIMO antenna...... 203

Fig. 6.6. Simulated reflection coefficients of the four different proposed antenna orientations...... 204

Fig. 6.7. Simulated transmission coefficients of the four different proposed antenna orientations...... 204

Fig. 6.8. Electromagnetic wave behaviour in different media [19]...... 206

Fig. 6.9. Simple two-port network...... 206

Fig. 6.10. The geometrical configuration of SRR unit cell...... 209

Fig. 6.11. The simulation setup of SRR unit cell...... 209

Fig. 6.12. The calculated permittivity of the proposed SRR unit cell...... 210

Fig. 6.13. Calculated permeability of the proposed SRR unit cell...... 210

Fig. 6.14. Transmission and reflection coefficients of the proposed SRR unit cell...... 211

Fig. 6. 15. Configurations of the proposed antenna; (a) Antenna without metamaterial resonators (b) Antenna with metamaterial resonators...... 211

Fig. 6.16. Simulated reflection coefficients with and without metamaterial resonators.

...... 213

Fig. 6.17. Simulated transmission coefficients with and without metamaterial resonators.

...... 213

xxi Fig. 6.18. Simulated reflection coefficients of the proposed MIMO antenna with a different number of metamaterial resonators...... 214

Fig. 6.19. Simulated transmission coefficients of the proposed MIMO antenna with a different number of metamaterial resonators...... 215

Fig. 6.20. Variation of calculated efficiency versus frequency for the single antenna,

MIMO antenna without decoupling and MIMO antenna with decoupling...... 216

Fig. 6.21. Simulated peak gain for the single antenna, MIMO antenna without decoupling and MIMO antenna with decoupling...... 217

Fig. 6.22. Equivalent circuit model of the proposed MIMO antenna...... 218

Fig. 6.23. Comparison of reflection coefficients (S11) for EM model and circuit model of the proposed MIMO antenna...... 219

Fig. 6. 24. Comparison of transmission coefficients (S21) for EM model and circuit model of the proposed MIMO antenna...... 219

Fig. 6.25. The prototype of the proposed antenna, (a) Front view; (b) Back view...... 221

Fig. 6 26. Simulated and measured reflection coefficient S11 of the proposed antenna.

...... 221

Fig. 6.27. Simulated and measured transmission coefficient S21 of the proposed antenna.

...... 222

Fig. 6.28. Simulated and measured ECC of the proposed MIMO antenna...... 223

Fig. 6.29. Simulated and measured capacity loss of the proposed MIMO antenna ...... 223

Fig. 6.30. The calculated values of TARC with different phases...... 224

Fig. 6.31. The calculated average value of TARC...... 225

Fig. 6.32. The simulated diversity gain of the proposed MIMO antenna...... 225

Fig. 6.33. Simulated and measured radiation patterns of the proposed antenna for two planes [(1 and 2) y-x plane and (3 and 4) z-x plane] at (a) 2.4 GHz (b) 3.4 GHz and (c)

xxii 3.6 GHz. Port 1 is excited and port 2 is terminated. Solid curve: simulated results, dashed curve: measured results, “1 and 3” are co-polar components and “2 and 4” are cross-polar components ...... 226

Fig. 6.34. The simulated 3-D pattern of the proposed MIMO antenna at (a) 2.4 GHz (b)

3.4 GHz and (c) 3.6 GHz...... 227

Fig.7.1. The proposed antenna geometry. (a) Front view; (b) Back view...... 237

Fig.7.2. Schematic structure of the proposed MIMO antenna showing DGS1; (a) Top layer, (b) Bottom layer...... 239

Fig.7.3. Simulated S-parameters of the proposed MIMO antenna with and without DGS1.

...... 239

Fig.7.4. Schematic structure of the proposed MIMO antenna showing DGS2; (a) Top layer, (b) Bottom layer...... 240

Fig.7.5. Simulated S-parameters of the proposed MIMO antenna with and without DGS2.

...... 240

Fig.7.6. Schematic structure of the proposed MIMO antenna showing DGS1 and DGS2;

(a) Top layer, (b) Bottom layer...... 241

Fig.7.7. Simulated S-parameters of the proposed MIMO antenna with DGS1 and DGS2.

...... 242

Fig.7.8. Schematic structure of the proposed MIMO antenna with NL and without DGS1 and DGS2; (a) Top layer, (b) Bottom layer...... 243

Fig.7.9. Simulated S-parameters of the proposed MIMO antenna without DGS1 and DGS2 and with NL...... 243

Fig.7.10. Schematic structure of the proposed MIMO antenna showing DGS1, DGS2 and

NL; (a) Top layer, (b) Bottom layer...... 245 xxiii Fig.7.11. Simulated S-parameters of the proposed MIMO antenna with DGS1, DGS2 and

NL...... 245

Fig.7.12. The simulated reflection coefficient of the proposed MIMO antenna with different lengths for the NL (L1=28 mm, L2=26 mm, L3=24 mm and L4=22 mm). .. 246

Fig.7.13. The simulated transmission coefficient of the proposed antenna with different lengths of the NL (L1=28 mm, L2=26 mm, L3=24 mm and L4=22 mm)...... 246

Fig.7.14. Surface current distribution at 2.4 GHz. (a) Without NL, (b) With NL...... 247

Fig.7.15. Simulated reflection coefficient S11 with and without neutralization line ..... 248

Fig.7.16. Simulated transmission coefficient S21 with and without neutralization line.

...... 249

Fig.7.17. Variation of calculated efficiency versus frequency for the proposed MIMO antenna with and without decoupling methods...... 250

Fig.7.18. Simulated peak gain for the proposed MIMO antenna with and without decoupling methods...... 250

Fig.7.19. The prototype of the proposed antenna, (a) Front view; (b) Back view...... 251

Fig.7.20. Comparison of reflection coefficient S11 using simulated and measured results.

...... 252

Fig.7.21. Comparison of transmission coefficient S21 using simulated and measured results...... 253

Fig.7.22. The measured envelope correlation coefficient...... 254

Fig.7.23. Simulated and measured channel capacity loss of the proposed antenna...... 254

Fig.7.24. Simulated and measured TARC of the proposed antenna...... 255

Fig.7.25. The simulated diversity gain of the proposed MIMO antenna...... 256

Fig.7.26. Simulated and measured radiation patterns of the proposed antenna for two planes [(1 and 2) y-x plane and (3 and 4) z-x plane] at (a) 2.4 GHz, (b) 3.5 GHz, (c) 5.5

xxiv GHz and (d) 5.8 GHz. Port 1 is excited and port 2 is terminated. Solid curves: simulated results, dashed curves: measured results, “1 and 3” are co-polar components and “2 and

4” are cross-polar components...... 257

Fig.7.27. The 3-D patterns of the proposed MIMO antenna at (a) 2.4 GHz, (b) 3.5 GHz,

xxv List of Tables

Table 2.1. Summarized of state of the art on the antenna placement and orientation method...... 31

Table 2.2. Summarized states of the art on the neutralization line method...... 38

Table 2.3. Summarized states of the art on the metamaterial method...... 43

Table 2.4. Summarized states of the art on the parasitic element method...... 50

Table 2.5. Summarized states of the art on the DGS method...... 57

Table 2.6. Comparison of mutual coupling reduction methods...... 58

Table 2.7. Previous work on narrow-band MIMO antennas...... 60

Table 2.8. Previous work on wideband MIMO antennas...... 62

Table 2.9. Previous work on multi-band antennas...... 64

Table 2.10. Summarized states of the art on the hybrid method element method...... 71

Table 3.1. Single antenna parameter values...... 107

Table 3.2. MIMO Antenna parameter values...... 113

Table 3.3. The effect of the position of the NL on the reflection coefficients...... 120

Table 3.4. The effect of the position of the NL on the transmission coefficients...... 120

Table 3.5. The effect of the width of the NL on the reflection coefficients...... 122

Table 3.6. The effect of the width of the NL on the transmission coefficients...... 122

Table 3.7. Comparison with other works...... 136

Table 4.1. Dimensions of the proposed MIMO antenna...... 145

Table 4.2. Comparison with other works...... 161

xxvi Table 5.1. The dimensions of the proposed MIMO antenna (mm)...... 167

Table 5.2 Comparison with other works...... 191

Table 6.1. The optimized dimensions of the proposed single antenna ...... 199

Table 6.2. The optimized dimensions of the proposed SRR unit cell...... 210

Table 6.3. The optimized dimensions of the proposed MIMO antenna...... 212

Table 6.4. Comparison with other works...... 228

Table 7.1 The characteristics of the proposed antenna at different frequencies...... 255

Table 7.2. Comparison with other works...... 259

xxvii List of Abbreviations

Abbreviation Name AI Artificial Intelligence CCL Channel Capacity Loss CLLs Capacitively Loaded Loops CSRR Complementary Split Ring Resonator DCS Digital Cellular System DG Diversity Gain DGS Defected Ground Plane ECC Envelope Correlation Coefficient EM Electromagnetic EFIE Electric Field Integral Equation IoT Internet of Things LTE Long Term Evolution MEG Mean Effective Gain MIMO Multiple Input Multiple Output MISO Multiple Input Single Output mmWave Millimetre Wave NL Neutralization Line PEC perfectly electrically conducting PMC perfectly magnetically conducting PIFA Planar Inverted F Antenna RL Return Loss SM Spatial Multiplexer SNR Signal To Noise Ratio SRR Split Ring Resonator STBC Space-Time Block Coding SIMO Single Input Multiple Output RF Radio Frequency TL Transmission Line TARC Total Active Reflection Coefficient UMTS Universal Mobile Telecommunications Service

xxviii UWB Ultra-Wideband V-BLAST Vertical-Bell Labs Layered Space-Time VNA Vector Network Analyser VR Virtual Reality VSWR Voltage Standing Wave Ratio WLAN Wireless Local Area Network WiMAX Worldwide Interoperability for Microwave Access Wi-Fi Wireless Fidelity XPR Cross Polarization power ratio

xxix List of Symbols

Notation Definition λo Wavelength in free space λg Guided Wavelength inside substrate Γ Reflection coefficient t 훤a Total active reflection coefficient 훤c Mean SNR for the diversity system 훾c Instantaneous SNR for the diversity system σ Conductivity (S/m) θ Phase of the input feeding 휓R Receive correlation coefficient matrix 휓T Transmit correlation coefficient matrix ηrad Radiation efficiency th Ai (Ω) Field pattern of the i elements C Channel capacity (bits/Hz/s) c Speed of light in free space = 2.99792458 × 108 (m/s) dB Decibel 휖푟 Dielectric constant (Dimensionless) 휖푟푒푓푓 Effective dielectric constant (Dimensionless) 휖표 Permittivity of free space = 8.854×10-12 (F/m) D Directivity fr Resonance frequency (Hz) f Operating frequency (Hz) G Gain Gθ (θ,∅) and Antenna gain components GØ (θ, ∅) GHz Gigahertz, 109 hertz h Thickness of the substrate (mm) h (Ω) Incoming waves H Normalized channel coefficients matrix HT Conjugate transpose of H matrix IR Identity matrix k Correction factor ko Free-space propagation constant n Refractive index Ni Receiver noise No power spectral density of noise N Number of antenna elements P Transmitted power Pθ (θ,∅) And Represent the channel model PØ (θ, ∅) Sij Scattering matrix elements ρe Envelope correlation coefficient tan δ Dielectric loss tangent

xxx Tj Transmitted Signals Ri Received signals μ Absolute permeability = μ0.μr (H/m μo Permeability of free space = 4π × 10-7 (H/m) W Bandwidth w width of the feed line z Relative impedance of the thin dielectric slab

xxxi CHAPTER ONE

Introduction

The development of wireless communication technology has raised demands for higher data rate and larger channel capacity in wireless applications such as Bluetooth, Wireless

Fidelity (Wi-Fi), Long Term Evolution (LTE), Wireless Local Area Network (WLAN) and Worldwide Interoperability for Microwave Access (WiMAX). Therefore, wireless communication technology is rapidly shifting from one antenna system to multiple antenna (MIMO) systems to fulfil the higher data rate and larger capacity requirements

[1, 2]. One of the main requirements for MIMO antenna systems to achieve maximum channel capacity is that the communication channels between the sender and receiver need to be totally uncorrelated. That means multiple radiators in transmitter and receiver ends should work individually. There are two main reasons that increase the channel correlation, the first one is the coupling inside the structure of the antenna. This parameter

(actually its inverse) is called antenna isolation and can be measured through the S- parameters of the antenna. The second factor is the characteristics of the radiation pattern of the antenna, which could increase the channel correlation even if the radiating elements are well isolated. This factor can be measured through the correlation coefficients of the antennas. In MIMO antenna systems, it is crucial to achieve high isolation and low correlation coefficient values to maximize the channel capacity. In an antenna array system, the coupling can be enhanced by increasing the physical distance between the radiating elements (at least λ/2). But this solution is inappropriate in contemporary handheld wireless devices with limited space availability. The deployment of MIMO

1 technology in mobile communication systems has led to application of multiple radiating antennas at the base station and at the mobile terminal as well. Generally, there is no firm limitation on space to design a MIMO antenna system operating at low frequencies (in

700-900MHz range) for the base station because the space between the radiating antennas can be increased to achieve high isolation.

The major challenge is to design mobile terminals operating in the sub-GHz frequency range with a high port-to-port isolation, due to the limited space available for the antenna system.

1.1 Motivation

As mentioned above, the requirement for multiple antennas in modern wireless systems is indisputable but combining these multiple radiating antennas in close proximity will lead to increases in the problems of isolation and channel correlation. The performance of the multiple-antenna system will be degraded by increasing the coupling between the antennas. In other words, it degrades the designed radiation efficiency and peak gain because some of the energy that is intended for one antenna will be collected by a neighbouring antenna [3, 4]. Another drawback associated with low isolation between the radiating elements is the high channel correlation which in turn will minimize the capacity of the MIMO system [5-7]. Consequently, it is essential to isolate the antennas to improve the performance of MIMO antenna systems, as well as to simplify the system design by creating near-ideal stand-alone operating conditions for individual radiating elements [8].

In this work, a comprehensive review of decoupling techniques that may be applied between the radiating elements of MIMO antenna systems in different wireless

2 applications is provided. This survey illustrates that there are different parameters that determine the choice of an appropriate decoupling technique. These parameters include the geometry of the antenna, the frequency of operation and the bandwidth of the antenna.

While previous researches have achieved excellent outcomes, most have had a few impediments. In other words, there were many gaps in previous researches in terms of adopting one coupling reduction scheme for multiband wireless applications. Therefore, the main motivation of this work is to overcome some of these impediments by designing a new low-profile MIMO antenna with integrated decoupling approaches for multiband applications, thereby improving MIMO antenna system capabilities.

1.2 Aim and Objectives

The main aim of the work reported in this thesis was to investigate the potential of integrating different decoupling techniques that can be used for MIMO antennas to improve the isolation of adjacent radiating elements within the structure to provide better diversity performance. Specifically, five decoupling techniques will be investigated.

These techniques include metamaterial resonators, antenna placement and orientation, parasitic elements, neutralization line (NL) and defected ground structure (DGS). Each technique is integrated with another technique/s and applied to a recently developed multi-band 2×2 MIMO antenna system to assess the effectiveness of integrating each pair of techniques on antenna performance.

A summary of the thesis objectives is listed below:

• Study the principle of operation of a MIMO antenna system and the parameters

affecting its performance, such as the scattering parameters, Envelope Correlation 3 Coefficients (ECC), Total Active Reflection Coefficient (TARC), Channel

Capacity Loss (CCL), Diversity Gain (DG), 2D and 3D antenna radiation patterns,

peak gain and antenna radiation efficiency.

• Investigate and verify the influences of integrating a DGS approach with a

neutralization line approach on three different proposed multi-band MIMO

antenna designs to enhance the isolation between the radiating elements.

• Design a low-profile printed MIMO antenna with two double concentric

rectangular rings. The benefit of using the self-similarity property is used to

achieve multi-band operation. In order to reduce mutual coupling, the DGS

decoupling approach is integrated with five parasitic elements in the bottom layer

of the MIMO antenna substrate.

• Design a new compact-size multi-band 2×2 MIMO antenna to work in WLAN

and Sub-6 GHz 5th generation (5G) bands. The antenna placement and orientation

technique is first implemented to minimize the coupling at the second frequency

band. Then, four split ring resonators (SRRs) are added between the radiating

elements to enhance the coupling at the first frequency band.

The above goals will be approached through theoretical and computer simulations using

CST Microwave Studio software (CST MWS). The corresponding results are to be validated by experimental measurements.

1.3 Original Contributions to Scientific Knowledge

The original contributions to scientific knowledge of this thesis are in the area of designing multi-band MIMO antennas with combined decoupling techniques to improve

4 the isolation between the radiating elements within the antenna structure and provide better diversity performance. Some of these contributions presented in this thesis are listed below:

• Designing a new low profile dual-band 2×2 MIMO antenna in the form of double

T-shaped monopole antennas with integrated decoupling techniques that

combines DGS and NL. This hybrid technique has helped to improve the coupling

of the two relevant bands and enhance system performance significantly.

• Designing a tri-band 2×2 MIMO antenna to fulfil the requirements of WLAN and

WiMAX applications. A new technique is presented to improve the isolation by

integrating DGS and NL decoupling techniques where the NL is placed in the

defected area to connect the ground planes. The main function of this method is

to reduce the coupling caused by the surface current in the ground plane.

• To improve the isolation between the radiating elements of a dual-band 2×2

MIMO antenna, a new hybrid decoupling approach that combines five parasitic

elements and DGS has been applied and implemented on the bottom layer of the

antenna structure.

• Designing a new 2×2 dual-band MIMO antenna consisting of two arc-shaped

elements for WLAN and Sub-6 GHz 5G applications. An integrated decoupling

approach that combines the antenna placement and orientation with metamaterial

resonators in the form of split-ring resonators (SRRs) are applied to suppress the

antenna coupling in the two frequency bands.

• A combination of three defected areas with a new NL configuration is used to

reduce the coupling and improve the impedance matching of two symmetrical G-

shaped slotted MIMO antennas working at WLAN and WiMAX bands.

5 1.4 Author’s Publications

Some of the important results obtained in this work have appeared in the following publications:

Journal Articles:

• Adham M. Saleh, Tariq A. Nagem, Raed A. Abd-Alhameed, James M. Noras

and Chan H. See, “Mutual Coupling Reduction of Dual-Band Uni-Planar MIMO

System Using Neutralization Line Technique,” Applied Computational

Electromagnetics Society, vol. 35, no. 2, pp. 176–186, February 2020.

• Adham M. Saleh, Tariq A. Nagem, Raed A. Abd-Alhameed and Khalil H.

Sayidmarie, “Dual-Band Printed MIMO Antenna Decoupling Based on a

Parasitic Elements and DGS Approaches.” Progress in Electromagnetics

Research, 2020. (Under Review).

• Adham M. Salah, Naem A. Jan, Tariq A. Nagem, Raed A. Abd-Alhameed, and

James M. Noras, “Isolation Enhancement of Compact Dual-Band Printed MIMO

Antenna Using Two Integrated Decoupling Techniques,” International Journal

of Electrical and Computer Engineering, 2020. (Under Review).

• Lashab, Mohamed, Chess‐Eddine Zebiri, Linda Djouablia, Mounir Belattar,

Adham Saleh, Fatiha Benabdelaziz, and Raed Abd‐Alhameed. “Characterization

of horn antenna loaded with CLL unit cell.” Microwave and Optical Technology

Letters, vol. 60, no. 8, pp. 1847-1856, 2018.

6 • Wan Noor N.W. Marzudi, Zuhairiah Z. Abidin, Siti Z. M. Adham M. Saleh,

Muji, Issa T. Elfergani, Raed A. Abd-Alhameed, James M. Noras. “Uni-Planer

MIMO Antenna for WLAN and WiMAX Wireless Services.” International

Journal of Computer and Information Technology, vol. 08, no. 03, pp. 78-83, May

2019.

Book chapter:

• See, Chan H., Adham M. Saleh, Ali A. Alabdullah, Khalid Hameed, Raed A.

Abd-Alhameed, S. M. R. Jones, and Asmaa H. Majeed. “Compact Wideband

Printed MIMO/Diversity Monopole Antenna for GSM/UMTS and LTE

Applications,” In Antenna Fundamentals for Legacy Mobile Applications and

Beyond, pp. 191-209. Springer, Cham, 2018.

International Conferences:

• Adham M. Saleh, Khalil H. Sayidmarie, Raed A. Abd-Alhameed, Stephen M.

R. Jones, James M. Noras, and Peter S. Excell. “Compact tri-band MIMO antenna

with high port isolation for WLAN and WiMAX applications,” In 2016

Loughborough Antennas & Propagation Conference (LAPC), pp. 1-4. IEEE,

2016.

• Adham M. Salah, Majid S. A. Alkhambashi, Naem A. Jan, Tariq A. Nagem,

Raed A. Abd-Alhameed and James M. Noras, “Isolation Enhancement of

Compact Dual-Band Printed MIMO Antenna for WLAN and 5G

7 Applications,” In 2020 IMDC-SDSP International Conference, Turkey, 2020.

(Accepted).

• Naem A. Jan, Adham M Saleh, Lashab M., Abdussalam F.M., Djouablia L.,

Abd-Alhameed R.A. (2017) “A Compact CSRR Loaded Monopole Antenna

with Defected Ground Structure for Mobile WLAN and WiMAX

Applications,” In: Otung I., Pillai P., Eleftherakis G., Giambene G. (eds)

Wireless and Satellite Systems. WiSATS 2016. Lecture Notes of the Institute

for Computer Sciences, Social Informatics and Telecommunications

Engineering, vol 186. Springer, Cham

• Lashab Mohamed, Naeem A. Jan, Adham M. Saleh, Ouchtati Salim, Chemss-

Eddine Zebiri, Djemili Rafik, Raed A. Abd-Alhameed, and Fatiha Benabdelaziz.

“Electrically small antenna loaded with CRLH for notched-band application,”

In 2017 Internet Technologies and Applications (ITA), pp. 305-308. IEEE, 2017.

• Ahmad, I et al. “Current technologies and location based services,” In 2017

Internet Technologies and Applications (ITA), pp. 299-304. IEEE, 2017.

1.5 Thesis Layout

This thesis constitutes a collection of the studies in the frame of the mentioned objectives.

The outcomes of the performed work are expressed in eight chapters. Beyond this introductory chapter, Chapter Two shows the main concept of MIMO technology, followed by its theoretical background and then a comprehensive review of the previous works, starting from the history of MIMO antenna systems and the main drawbacks

8 associated with MIMO antennas. After that, various decoupling techniques commonly employed in the literature are presented. A further discussion of key performance metrics used to measure the performance of MIMO systems is also presented.

Chapter Three presents a compact dual-band MIMO antenna operating at WLAN bands

(2.4/5.2/5.8 GHz). The schematic structure of this antenna consists of two double T- shaped radiating elements with integrated DGS and neutralization line approaches. The optimum value of isolation is obtained by optimizing the dimensions of the neutralization line and DGS.

Chapter Four investigates a low-profile tri-band MIMO antenna operating in WLAN

(2.4/5.2/5.8 GHz) and WiMAX (3.5 GHz) bands. This antenna design employs two printed uni-planar monopole antennas. A new technique combining the DGS approach with the neutralization line approach is presented. The main benefit of these techniques is to reduce the coupling caused by the surface current in the ground plane.

Chapter Five focuses on the effect of integrating the DGS decoupling approach with parasitic elements to mitigate mutual coupling. This hybrid decoupling approach is applied on a MIMO antenna structure that contains two radiating elements in the form of a double square ring to operate in two different frequency bands.

Chapter Six presents a dual-band MIMO antenna consisting of two arc-shaped elements, intended for WLAN and Sub-6 GHz 5G applications. The antenna placement and orientation technique is used to enhance the isolation in the Sub-6 GHz 5G band while

9 metamaterial resonators in the form of split-ring resonators are introduced to reduce the antenna coupling in the WLAN band.

Chapter Seven presents a low-profile dual-wideband MIMO antenna for WLAN and

WiMAX services. The optimised structure consists of two symmetrical G-shaped slotted patch antennas. A combination of neutralization line and defected ground plane techniques is used to reduce the coupling. The antenna impedance matching is improved by printing stubs on the feeding lines near to radiating elements.

Chapter Eight summarizes the conclusions from the previous chapters and presents the guidelines for future work as deduced from these conclusions.

10 References

[1] R. G. Vaughan and J. B. Andersen, “Antenna diversity in mobile communications,”

in IEEE Transactions on Vehicular Technology, vol. 36, no. 4, pp. 149-172, Nov.

1987.

[2] G. J. Foschini and M. J. Gans, “On limits of wireless communications in a fading

environment when using multiple antennas,” Wireless Personal Communication, vol.

6, no. 3, pp. 311–335, Mar. 1998.

[3] R. Vaughan and J. B. Andersen, “Channels, Propagation and Antennas for Mobile

Communications,” Institution of Engineering and Technology, 2003.

[4] M. R. Sharawi, “Printed MIMO antenna engineering,” Norwood: Artech House,

2014.

[5] P. Fletcher, M. Dean, and A. Nix, “Mutual coupling in multi-element array antennas

and its influence on MIMO channel capacity,” Electronics Letters, vol. 39, pp. 342-

344, 2003.

[6] M. Ozdemir, E. Arvas, and H. Arslan, “Dynamics of spatial correlation and

implications on MIMO systems,” IEEE Communications Magazine, vol. 42, pp. S14

-S19, 2004.

[7] R. Janaswamy, “Effect of element mutual coupling on the capacity of fixed length

linear arrays,” IEEE Antennas and Wireless Propagation Letters, vol. 1, no. 1, pp.

157-160, 2002.

[8] A. H. R. Thuwaini, “Mutual coupling suppression in multiple microstrip antennas for

wireless applications,” PhD dissertation, Brunel University London, 2018.

11 CHAPTER TWO

Background and Literature Review of MIMO Antenna Systems

Recently, wireless communication systems are required by modern applications to achieve higher and higher data rates with maximum channel capacity. Since increasing the bandwidth is costly and there is some limitation to use higher-order modulation types, a new approach that takes the benefits of the transmission channel have to be adopted.

MIMO technology provides a significant improvement to data rate and channel capacity.

The main feature of MIMO systems is that it utilizes the multipath propagation, which is a pitfall of wireless communication systems, and turns it into an advantage for the user.

On the other hand, MIMO effectively improves spectral efficiency by increasing the capacity of the wireless communication system without any increase in available bandwidth. This chapter gives an introduction to the concept of MIMO technology and terminology and explains its theoretical background. It also presents a comprehensive survey of the literature available regarding mutual coupling reduction methods of MIMO antennas by showing the main features of each method and its cons and pros. This survey helps to find the gaps in knowledge that drive this work. The main diversity metrics of

MIMO antenna systems that use to describe MIMO antenna performance are presented in detail at the end of this chapter.

2.1 The Concept of MIMO Technology

MIMO technology can be defined as an advanced technique that utilises multiple transmit and receive antennas for multiplying the capacity of a wireless communication system

12 within a fixed bandwidth. In other words, MIMO technology exploits a solution in which a communication system can send and receive multiple independent channels simultaneously through the same wireless link (and same frequency band) using multiple antenna topologies without any extra radiation power consumption. It exploits the additional paths provided in a rich scattering environment [1, 2]. MIMO technology has also been distinguished as a backbone for next generation wireless communication technology due to its ability to enhance system reliability and improve channel capacity by applying multiple antennas. Initially, MIMO technology was presented as a practical technique that can improve the limitation in the data rate experienced by single input single output (SISO) systems. Moreover, different networks can apply MIMO technology to enhance the capacity of the channel, the reliability of the system and the transmission speed of the data [3] by exploiting the highest capacity of a wireless communication system [4].

There are related technologies that may be used in wireless communication systems such as Single Input Multiple Output (SIMO), Multiple Input Single Output (MISO) and

Multiple Input Multiple Output (MIMO). In SIMO, there is one transmitter antenna and multiple receiver antennas. This configuration is known as receiver diversity. This technique enables the receiver to combat fading by receiving multiple signals coming from different independent channels conveying the same information. In MISO, there are multiple transmitters and one single receiver: redundant data is transmitted through multiple transmitter antennas. Thus, the receiver has a greater probability to receive the transmitted data correctly and hence reduce the effects of multipath signal propagation

[5]. MIMO technology uses multiple antennas at the transmitter and multiple antennas at the receiver. Therefore, the transmitter can send multiple data streams in the environment

13 of multipath. Then, the receiver end will receive more than one channel of data due to the multipath environment. After that, the received data should be analysed using a special

MIMO algorithm in order to recover the multiple transmitted data streams [5]. Fig. 2.1 shows the configuration of different multiple antennas systems.

Fig. 2.1. Configuration of different multiple antenna systems.

2.2 Theoretical Background of MIMO Technology

As mentioned earlier in this chapter, a MIMO system consists of multiple antennas at the transmitter and receiver ends respectively. The multiple antennas at the transmitter will help to send multiple data streams and each antenna in the receiver end will receive multiple data copies. Fig. 2.2 shows the block diagram of a MIMO system. Assuming a

MIMO system with N×M antennas, the data will be divided into N parallel streams. Then,

14 the parallel streams will be received by the M receiving antennas. The formula of the received signal at the receiver end can then be written as follows [6]:

R1=(H11×T1+ H12×T2+… ...... H1N×TN)+N1 R2=(H21×T1+ H22×T2+… ...... H2N×TN) +N2 . (2.1) . .

RM=(HM1×T1+ HM2×T2+… ...... HMN×TN) +NM

th th where Ri and Tj are the received and transmitted signals by the i and j antennas respectively and the channel between them is represented by Hij while Ni is the receiver noise [7]. The transmitted signal and the received signal are related to each other. This relation can be explained in the following formula:

[푅]푇 = [퐻][푇]푇 +[N] (2.2)

The recovery of the transmitted signal can be evaluated by:

[푇]푇 = [퐻] ―1[푅]푇 ― [퐻] ―1N (2.3)

Fig.2.2. MIMO System block diagram.

From these formulas, it can be seen that spatial diversity and spatial multiplexing are used by a MIMO system to improve the capacity and/or quality of the link. The process in which the MIMO technology could increase the channel capacity depends on the Shannon channel capacity equation [7]:

푝 퐶 = 푊 log2 1 + (2.4) ( 푁표푊 )

Where:

C: is the channel capacity in bits/Hz/s.

W: is the bandwidth.

P: is the transmitted power.

N0: is the power spectral density of noise.

Signal to noise ratio (SNR) is represented in the equation by the term (P/WN0). Therefore, the channel capacity of the system can be increased by either raising the SNR or increasing the bandwidth of the channel. As is well known, the signal to noise ratio has a logarithmic relation with the channel capacity, which means there is little effect from increasing the signal to noise ratio above a specific value: this is because the channel capacity will be in the saturation mode. Consequently, it is not efficient to increase the transmitted energy and this could cause a conflict with other communication channels

[8]. The second factor that could be used to improve channel capacity is the bandwidth.

There are limitations on increasing the bandwidth linearly in a communication system due to the high cost of acquiring new spectrum, the severe fading in a multipath

16 environment, and the restricted transmitted energy to avoid interference with other channels [9]. Therefore, the MIMO technology has been used to counter the previous limitations through improving the channel capacity of the communication system within a specific transmitted energy and fixed bandwidth. Finally, MIMO antenna systems can be considered as a key element of most modern emerging wireless technologies.

2.3 History of MIMO Antenna Systems

The multiple antenna concept was first conceived during early 1998 by pioneer Foschini, where multiple antennas at the transmitter and receiver in a communications system have been employed to enhance the channel capacity [10]. During 1998-1999, a very clever concept of “Spatial multiplexing” also known as V-BLAST (Vertical-Bell Labs Layered

Space-Time) communication technique, as shown in Fig. 2.3, was developed by

Wolniansky, Golden and Foschini for enhancing spectral efficiencies [11-13]. In this technique, a high data rate stream is separated into multiple low rate sequences. These low rate sequences are transmitted simultaneously using multiple antennas with the same frequency of operation. In the receiver end, the low rate sequences are combined from different paths to reconstruct the actual transmitted data.

On the other hand, the spatial multiplexing technique does not help in improving the diversity gain; the latter can be considered as the main parameter for minimizing the fading in any communication system. Therefore, Zheng, Zheng and Tse [14] in 2003 illustrated that there is a trade-off between diversity gain and spatial multiplexing gain in a communication channel in order to achieve the optimum performance of the system. In

2005, Walter et al. [15] explained that the combination of the spatial multiplexer (SM)

17 and Space-Time Block Coding (STBC) technologies will lead to improvement in the diversity of the MIMO system and increase the average data rates in the MIMO channel.

Fig. 2.3. Configurations of V-BLAST System [11].

2.4 Antennas for MIMO Systems

In the last few years, the size and weight of mobile terminals have been reduced dramatically. The rapid evolution of the antennas used for mobile terminals has been triggered by this reduction in the mobile terminal’s size. Moreover, the demand for higher data rates and larger channel capacity has helped to evolve the antenna requirements for a handset from one single radiator antenna to multi-radiator solutions. As a result, the antennas for compact handsets are becoming more difficult to design. The antennas are requested to be compact whilst their performances have to be preserved. Nevertheless, the gain and bandwidth are usually degraded by reducing the antenna’s size. In MIMO technology, multiple antennas will be integrated in a mobile handset to achieve the benefit of the diversity that these antennas provide. Consequently, it is more challenging to design a small mobile terminal with multiple antennas for MIMO systems, compared to a mobile terminal with one single conventional antenna. The major difficulty in MIMO

18 systems for compact terminals is the limited space between the radiating elements because when multiple antennas are placed in close proximity, mutual coupling can occur between them [6]. This degrades the isolation between the radiating elements and no diversity gain can be achieved. On the other hand, the efficiencies of the radiating elements will also be decreased by the existence of the coupling between them because some of the power that should be radiated will be collected by another radiating element.

In diversity and MIMO systems, sufficient isolation between the radiating elements is crucial. Different techniques have to be developed to minimize the mutual coupling and enhance the isolation between the radiating elements. One of the straightforward methods is by placing the two radiating elements far apart from each other, at least a half wavelength or more at the operating frequency. Furthermore, the size of the antenna should also be taken into consideration because the frequency of operation is inversely proportional to the dimensions of the antenna.

As shown in previous sections, increasing the number of antenna elements of the MIMO system will lead to a linear increase in the capacity of the MIMO system. On the other hand, practical investigations show that the corresponding capacity of the system may be minimized if there is a high correlation in the received signals in any of the various antenna elements [16]. The Envelope Correlation Coefficient (ECC) can be considered as a practical method to calculate and characterize the diversity capability of a MIMO antenna system. A better MIMO system performance can be achieved with lower values of correlation coefficients [6].

19 Furthermore, another factor that plays a significant role in determining the MIMO antenna performance is the balanced branch power-mean effective gain (MEG). The system performance can suffer from diversity loss as a result of an imbalanced power in the diversity branches [17]. The total efficiency of the antenna influences this imbalance.

However, since practical considerations lead to an expectation that the diversity can be applied under any possible circumstances, the antenna-channel mismatch is also quite important. Consequently, the mean effective gain is widely used, due to its unique feature of representing all the influences of the total efficiency of the antenna, the gain of the antenna and the environment of the wireless system [18]. In order to obtain an optimal diversity performance, the multiple antenna system has to fulfil the balanced power requirement [6]:

MEGantenna1≈ MEGantenna2 (2.5)

2.5 Mutual Coupling

The performance of the MIMO system suffers from another adverse factor which is the mutual coupling between the antennas. This factor will also limit the channel capacity of the system. The main reason for the coupling in MIMO antenna systems is the limited space between the antennas, hence the electromagnetic radiation from MIMO antenna elements will affect each other. This issue has greater effect with microstrip antennas due to the existence of surface waves in the substrate [19]. Therefore, the performance of the

MIMO system should be enhanced by minimizing the coupling between the antennas.

Allen and Diamond [20] studied the mutual coupling in array antennas in 1966. Their work has had a significant effect on much subsequent research.

20 In 1982, Malkomes [21] worked on the mutual coupling of the microstrip patch antenna array. The mutual coupling between the elements of a microstrip array were also analyzed by Penard and Daniel in 1982 [22]. The transmission line method was used by Van Lil et al. [23] in order to calculate the mutual coupling between the elements in a microstrip array in 1984. In 1997, Balanis [24] presented the most important causes of coupling between the elements in an antenna array, for example, the geometry and the type of the antenna, the element positions in antenna arrays and the feeding technique of the array elements. Ramirez and Flaviis [25] in 2000 studied the mutual coupling in a microstrip antenna array with linear polarization. They observed that when the adjacent elements transmit or receive in orthogonal senses of polarization the mutual coupling can be reduced. This array can be used in Bell Laboratories Layered Space-Time (BLAST) wireless communications systems.

In 2001, Svantesson and Ranheim [26] analyzed the performance of the elements of the antenna array for a mobile terminal in term of channel capacity and how this could be affected by mutual coupling. In 2002, Wallace and Jensen [27] studied the influence of mutual coupling on channel capacity. Yuehe Ge et al. [28] in 2004 analyzed the effect of changing the orientation of the antennas in order to minimize the coupling between them.

In 2005, the Electric Field Integral Equation (EFIE) was used by Mbonjo et al. [29] to investigate the impact of mutual coupling on the channel capacity of MIMO systems. Hui

[30] in 2007 explained several techniques that could be used for decoupling the elements in antenna arrays. Ali et al. [31] in 2007 studied a variety of methods that could be used to improve the isolation between the elements in an antenna array.

21 2.6 Mutual Coupling Reduction Methods

As explained earlier, mutual coupling in MIMO systems can be classified as the main reason for limitation of the channel capacity and diversity performance. For this reason, various techniques available in the literature have been developed in order to minimize the mutual coupling and enhance the performance of the system. For example, antenna configuration, decoupling networks, parasitic elements, Defected Ground Plane

Structures, neutralization lines, T-shaped junctions, etching slots on the ground plane and metamaterials [32]. This section will discuss several different methods that are explored in this thesis.

2.6.1 Antenna Placement and Orientation Method

The method of using antenna element placement and angular variation is commonly applied to minimize mutual coupling between the closely packed radiating elements. In this technique, the variation of space between the radiators plays a major role to achieve decorrelation or high isolation. It is easy to achieve mutual coupling less than -20 dB if the space between the radiators is greater than or equal to λ/2, where λ is free space wavelength at the centre frequency [33]. However, the distance between the radiators becomes smaller than λ/2 in the case of compact antenna systems for handheld devices.

Consequently, the mutual coupling effect needs to be taken into consideration. Therefore, placing the radiating elements at different orientations with respect to each other will help to enhance the isolation by taking advantage of the diversity in polarization [34].

In 2004, Carrasco et al. [33] studied the influence of the location of planar inverted-F antennas (PIFA) and monopoles. The PIFAs were studied under different arrangements.

22 The first case was collinear arrangement, where the two PIFAs are placed along a straight line passing through the feeding points. In this arrangement, minimum coupling was achieved by placing the two open ends of the PIFAs at the opposite sides of each other

(maximum space between them). The same procedure was applied in the next arrangement (parallel arrangement) where the PIFAs were set next to the parallel edges.

In this case, the maximum distance between the two ends produced maximum isolation between the PIFAs. In the last arrangement (orthogonal arrangement), the axes of the radiators were placed at 90o angles from each other. In this configuration, the Coupling is much larger when the open end of one antenna faces the feed point of the other.

Kin-Lu Wong et al. [35] in 2005 presented the effect of changing the position of the printed antenna within the board. In this research, an antenna system consists of two radiating elements, the first one working in the GSM/DCS band while the second one operated in the WLAN band. The position of the WLAN radiating element was changed along the circumference of the antenna system board and the mutual coupling between the two radiating elements was evaluated. It was observed that minimum coupling can be achieved when the two radiating elements were placed along the diagonal of the antenna system board as it provides maximum separation between the radiating elements.

In 2007, Chae et al. [36] investigated two antenna arrays in handsets by taking into consideration mutual coupling, correlation coefficient, and total active reflection coefficient (TARC). The outcomes illustrated that these three metrics are not only affected by antenna spacing: they were also directly influenced by radiation patterns. For this reason, it is preferable to recognize pattern diversity instead of spatial diversity to

23 minimize the correlation in compact antenna arrays. The configurations of three different cases are shown in Fig. 2.4. In 2010, Park and Jung [37] presented a compact two-channel

MIMO antenna consisting of two folded monopoles with low mutual coupling characteristics. The directions of different main beams were studied by placing the two folded monopoles in orthogonal arrangements at the edge of the ground plane. The geometry of the design is shown in Fig. 2.5.

Fig. 2.4. Three different configurations of the proposed dual array. [36].

Fig. 2.5. The geometry of the two folded antennas [37].

In 2010, Zhang et al. [38] investigated four radiating elements in the form of quarter- wavelength slots for a MIMO antenna. The low mutual coupling and correlation coefficient were obtained by arranging the radiating elements in an appropriate orientation. Slots were introduced in the ground plane for mutual coupling suppression purposes. The schematic structure of the proposed MIMO antenna is presented in Fig.

2.6. Al-Nuaimi and Whittow [39] in 2011 studied the effects of the position of the radiating elements on a MIMO antenna array of inverted-F antennas (IFA). Five different arrangements were investigated in this work. It was concluded that not only the space between the radiating elements will affect the mutual coupling but the antenna angular orientation as well. The five different arrangements are illustrated in Fig. 2.7. In 2012

Xiong et al. [40] designed a MIMO antenna consisting of a modified multiply-fed shielded loop antenna and an E-dipole. In this design, an orthogonal polarization arrangement was used to achieve high isolation and low envelope correlation coefficient

(ECC). The MIMO antenna presented there can be used as a reference antenna for upcoming collocated polarization-diversity-based MIMO antennas. A photograph of the fabricated antenna is shown in Fig. 2.8.

Fig. 2.6. The schematic structure of the four-element MIMO antenna with different element arrangements; (a) Front side, (b) bottom side [38].

Fig. 2.7. The five different cases of an IFA array [39].

Fig. 2.8. Photograph of the modified multiply-fed shielded loop antenna with an E-dipole

[40].

Lin and Liu [41] in 2013 investigated the spatial orientation and its ability to control mutual coupling and improve diversity metrics in a MIMO antenna system. This investigation has been achieved by examining different arrangements such as diagonally opposed placement, face to face and180o phase shift. The measured S-parameters of the three cases are shown in Fig. 2.9. In 2014, Jeong et al [42] studied the impact of changing the spatial orientation on the direction of the current flow, thus controlling the correlation factor. They designed a MIMO antenna that works at the WLAN/WiMAX/LTE bands with different configurations for the desired frequency of operation, as illustrated in Fig.

2.10.

Fig. 2.9. The measured S-parameters of the fabricated MIMO antenna in [41]; (a) Case

1, (b) Case 2, and (c) Case 3.

Fig. 2.10. The MIMO antenna proposed in [42]; (a) The geometry of the MIMO antenna

(b) The geometry of the single radiator.

In 2015, Khan and Sharawi [43] designed a four-element dual-band MIMO antenna system made up of microstrip annular slots. The lower frequency band was controlled by two elements while the upper frequency band was controlled by the other two elements.

Good isolation between the compact radiating elements was achieved with the help of antenna element placement and feeding methods. The fabricated design is shown in Fig.

2.11. Malviya et al. [44] in 2016 presented a compact four-element dual-band MIMO antenna. In this design, the radiating elements were orthogonally arranged to form the antenna structure. The isolation between the elements was improved by the orthogonally arranged ground for each element, as shown in Fig. 2.12. The measured mutual coupling in each intended frequency band was more than 21 dB.

Fig. 2.11. Photograph of the fabricated microstrip annular slots MIMO antenna; (a) Front side, (b) Bottom side [43].

Fig. 2.12. The prototype of the 4-element orthogonally arranged MIMO antenna; (a) Front side, (b) Bottom side [44].

Mathurand and Dwari [45] in 2018 designed a compact four-element ultra-wideband

(UWB) MIMO/diversity antenna. The UWB characteristic was obtained through identical monopole elements with three annular rings as shown in Fig. 2.13. Placing the radiating monopoles orthogonal to each other led to improved inter-element isolation of more than 15 dB. In 2019, Aw et al. [46] studied a 2×2 wide-band MIMO antenna with a bandwidth ranging from 4.85 to 6.25 GHz. In this design, the effect of keeping one of the

29 radiating elements in the same position while changing the location of the second element was investigated. Six different configurations were presented to achieve the highest isolation between the elements, as shown in Fig. 2.14.

Fig. 2.13. The proposed four-element UWB MIMO antenna with different arrangements;

(a) Schematic structure, (b) Fabricated antenna [45].

Fig. 2.14. The proposed MIMO antenna showing the six different configurations with their corresponding S-parameters [46].

30 Finally, a summary of state-of-art antenna placement and orientation methods, covering the important previous works is given in Table 2.1, which details the centre frequency, isolation, electrical size, antenna peak gain, antenna efficiency and ECC. In [43], higher efficiency of more than 84% and simple geometry with dual-band characteristics is achieved. Moreover, it has low mutual coupling of more than 16 dB and 21 dB at the two frequency bands respectively. Therefore, the work in [43] can be represented as a better performance compared to other designs illustrated in this section.

Table 2.1. Summarized of state of the art on the antenna placement and orientation method.

Centre Peak Antenna Isolation Ref. Frequency Electrical Size Gain Efficiency ECC Remarks (dB) (GHz) (dBi) (%) 0.825 33.6 Compact size and no

[35] 1.9 27.9 0.19λo ×0.275λo ×0.002λo gain, efficiency and ECC 2.5 33 Simple structure and no [36] 2.65 <17 0.39λo ×0.7λo×0.035λo gain, efficiency and ECC High efficiency and no [37] 2.475 <25 0.495λo ×0.495λo × 2.1 0.14 efficiency and substrate thickness High isolation and gain [38] 1.86 >30 0.43λo× 0.43λo×0.0094λo 5 and no efficiency and ECC Simple geometry and size

[39] 3 <30 1λo ×0.5λo ×0.008λo and no gain, efficiency and ECC High isolation, no -2 [40] 0.86 <40 0.49λo ×0.2λo × >1.15 <10 efficiency and substrate thickness Big size, simple structure [41] 2.595 >20 1.73λo ×1.47λo ×0.008λo <3.1 and no efficiency and ECC

2.4 -3 High gain, low ECC and [42] >20 0.8λo ×0.4λo ×0.008λo <10 5.5 no efficiency and gain Simple structure, good 2.575 >16 3.46 [43] 0.5λo ×0.5λo ×0.012λo >84 isolation and gain, high 5.25 >21 3.82 efficiency and no ECC

2.5 -2 Good isolation, low ECC [44] >17 0.58λo ×0.58λo ×0.012λo <3.6 <10 5.19 and no efficiency Simple structure, big size [45] 7.1 >15 0.85λo ×0.85λo ×0.037λo <5.7 >70 <0.05 and good isolation

-3 High isolation, big size [46] 5.35 >35 0.98λo ×0.35λo ×0.028λo <4 <10 and no efficiency

31 2.6.2 Neutralization Line Method

A neutralization line is one of the methods that can be applied for improving the performance of the MIMO system by decreasing the coupling between the antennas. This can be done by inserting a piece of stripline between the radiating elements. Thus, the current at a specific location on the radiating element will be sampled and its phase will be inverted in order to cancel the current from the adjacent radiating element.

In 2006, Diallo et al. [47] studied the reduction of the mutual coupling between two mobile phone PIFAs operating in the DCS1800 and UMTS Bands. They used a neutralization line method for connecting specific points on the radiator’s surface to enhance the isolation between them, as shown in Fig. 2.15. Ranvier et al. [48] in 2007 attempted to enhance the channel capacity by decreasing the mutual coupling and the efficiency of the system. They designed two rectangular patch antennas operating at 5.75

GHz and then used a neutralization line to connect the two elements as illustrated in Fig.

2.16. In 2010, Li, Han, Zhao, and Choi [49] designed two monopole radiating elements to form a MIMO antenna for wireless USB dongle application at WLAN band; they enhanced the isolation between the elements by inserting a NL between them as shown in Fig. 2.17. Su, Lee, and Chang [50] in 2012 designed two monopole MIMO antennas to cover the 2.4 GHz WLAN band. In this work, the neutralization line was added to connect the two radiating elements near to the feeding point, as shown in Fig. 2.18.

Fig. 2.15. Configurations of the DCS and UMTS PIFAs [47].

Fig. 2.16. The prototype of the proposed rectangular patch MIMO antenna operating at

5.75 GHz with NL [48].

Fig. 2.17. Proposed MIMO antenna with NL for wireless USB dongle application[49];

(a) Geometry of the proposed antenna, (b) Prototype of the proposed antenna.

Fig. 2.18. A prototype of the proposed MIMO antenna with NL near to the feeding point

[50].

In 2014, Elkhazmi et al. [51] designed two crescent-shaped radiating elements with a zigzag neutralization line, as shown in Fig. 2.19, to achieve good isolation and correlation coefficients. The frequency band was from 2.4 to 4.2 GHz to cover WLAN and WiMAX applications. Zhang and Pedersen [52] in 2016 designed a wideband neutralization line to connect ultra-wideband (UWB) MIMO antenna elements and reduce the coupling between them. This UWB MIMO antenna covers the band 3.1–5 GHz with high isolation between the elements. Fig. 2.20 illustrates the prototype of the proposed antenna with wideband neutralization line. In 2016, Shin, Kibria and Islam [53] implemented a folded- shaped monopole MIMO antenna with a neutralization line and a rectangular patch to achieve 10.0 dB of diversity gain (DG) with isolation greater than 14.0 dB. This MIMO antenna was designed to cover LTE, GSM, and UMTS band applications.

Fig. 2.19. Configurations of the crescent-shaped MIMO antenna with zigzag NL [51]; (a)

Top view, (b) Bottom view.

Fig. 2.20. Printed MIMO antenna with wideband NL [52]; (a) Top view, (b) Bottom view.

In 2018, Banerjee, Karmakar and Ghatak [54] exhibited a compact printed CPW-fed dual- band notched UWB MIMO antenna. A NL in the form of a modified 2nd order Hilbert fractal shape was implemented beside a modified circular defect in the common ground plane. Further, an isolation reinforcement for the MIMO system was also achieved by using a modified Minkowski boundary fractal geometry on the top edges of the respective

35 radiating elements. The design configuration of their proposed antenna is shown in Fig.

2.21. Chou et al. [55] in 2018 presented a compact dual band MIMO antenna with a decoupling approach for a tablet computer with full-metallic bottom cover. In this model, a NL alongside two series lumped capacitors were added to the antenna structure to suppress the coupling between the radiating elements, as shown in Fig. 2.22. Tiwari et al.

[56] in 2019 investigated three different cases of neutralization line shapes by adding them between the radiating elements of a compact 2×2 UWB MIMO antenna. The optimization process illustrated that adding a rectangular strip in the middle part of the neutralization line produced minimum mutual coupling. The measured bandwidth of 2×2 and 4×4 MIMO antennas offered 95.22% in frequency band 3.51–9.89 GHz and 96.47% in frequency band 3.52–10.08 GHz with isolations more than 24 dB and 23 dB, respectively. The three different cases of NL shapes are shown in Fig. 2.23.

Fig. 2.21. Design configuration of the proposed antenna with a modified 2nd order Hilbert fractal shaped NL [54].

Fig. 2.22. Schematic structure of the proposed MIMO antenna with NL and two series lumped capacitors for tablet computer [55].

Fig. 2.23. The geometry of the proposed MIMO antenna showing the three different cases of NL shapes [56].

Finally, a summary of the state-of-art of neutralization line methods, containing the important previous works is given in Table 2.2, which details the centre frequency, isolation, electrical size, separation distance between the radiating elements, antenna peak gain, antenna efficiency and ECC. In [52], a wideband neutralization line is presented with minimum substrate thickness and acceptable gain values. Further, the geometry of the antenna is simple compared to other antennas presented in this section. High efficiency and isolation values are also achieved.

37 Table 2.2. Summarized states of the art on the neutralization line method.

Centre Peak Antenna Isolation Separation Ref. Frequency Electrical Size Gain Efficiency ECC Remarks (dB) distance (GHz) (dBi) (%)

1.8 0.6λo×0.24λo 1.9 85 High efficiency and [47] >15 0.108λo 2.1 ×0.018λo 1.1 93 no ECC

Narrow bandwidth for isolation (0.1

[48] 5.75 >15 0.4λo GHZ) with S21 less than 15 dB while S11 BW is 0.5 GHz Very low simulated 0.12λo×0.33λo isolation barely [49] 2.49 15 0.04λo 2 40 0.02 ×0.006λo reach 15 dB and low efficiency 0.24λo×0.52λo High isolation and [50] 2.44 22 0.11λo 2 75 0.01 ×0.008λo low ECC 0.99λo×0.44λo Big size and [51] 3.3 >17 0.2λo 2.5 <0.01 ×0.008λo expensive substrate 0.47λo×0.44λo High efficiency and [52] 4.05 >22 0.03λo 2.8 90 <0.1 ×0.01λo isolation 0.793 12 0.5 66 0.26λo×0.1λo Low isolation at the [53] 1.956 12 2.04 62 <0.3 2.514 17 ×0.004λo 0.02λo 1.9 65 first two bands

0.73λo×0.9λo High isolation and [54] 8.275 >20 0.014λo 2 ≤0.1 ×0.044λo Complex geometry 2.442 16 0.49λo×0.06λo High isolation and [55] 0.14λo 5 >70 5.5 23 ×0.036λo large thickness 0.46λo×0.75λo Large size and high [56] 6.65 >22 0.09λo 4.3 78 <0.01 ×0.035λo isolation

2.6.3 Metamaterials Method

Metamaterials can be defined as artificial materials which can be characterized by their periodicity. Metamaterials have many features that could be used in the field of antennas.

One of these features is the ability to provide a stop band at a certain frequency. MIMO antenna systems have used this feature to minimize the coupling between the radiating elements. Many examples of using metamaterials to improve the isolation have been listed in the literature.

Hsu et al. [57] in 2009 used different configurations of capacitively-loaded loops (CLLs) to reduce the coupling between the radiating monopoles in a MIMO system for portable applications, as shown in Fig. 2.24. Lee, et al. [58] in 2011 designed rectangular radiating elements with L-shaped slots. Metamaterials in the form of split-ring resonators (SRRs)

38 were etched on the substrate in order to minimize the coupling between the elements, as illustrated in Fig. 2.25. This antenna was designed to cover the WiMAX frequency band.

Fig. 2.24. Configuration of the proposed MIMO antenna with metamaterials in the form of capacitively-loaded loops [57].

Fig. 2.25. Configurations of the proposed L-shaped slots MIMO antenna; (a) Top view,

(b) Bottom view, (c) Side view, (d) Radiating element, (e) SRR unit cell structure [58].

In 2011, Lee and Lee [59] used metamaterials to minimize the coupling between two loop radiators: more than 10 dB isolation has been achieved. Fig. 2.26 shows the configuration

39 of their proposed MIMO antenna. Lee, Ga and Choi [60] in 2012 designed an antenna system consisting of three layers and SRR. The results show that a good reduction in mutual coupling has been achieved with a frequency band from 2.26 GHz to 2.42 GHz.

Fig. 2.27 shows the fabricated MIMO antenna with SRR.

Fig. 2.26. Configurations of the proposed two loop radiators with metamaterial unit cells

[59]; (a) Top view, (b) Bottom view.

Fig. 2.27. Printed MIMO antenna with SRR [60]; (a) Top view, (b) Bottom view.

In 2013, Ketzaki and Yioultsis [61] used the concept of the metamaterial to design planar

MIMO monopole antennas with high isolation between them. The metamaterials have

40 the form of rectangular loop resonators inserted between the antenna elements. Different variations have been applied to optimize the performance of the MIMO antenna, as shown in Fig. 2.28. Zhai, Chen and Qing [62] in 2015 proposed a double-layer “mushroom” structure to reduce the coupling between four radiating elements of the MIMO antenna system, as shown in Fig. 2.29. In 2017, Farahani et al. [63] used a metamaterial polarization-rotator wall to suppress the coupling between millimetre-wave dielectric resonator antennas. The isolation was enhanced by 16 dB on average. Fig. 2.30 illustrates the construction of the proposed MIMO antenna.

Fig. 2.28. Configuration of the proposed MIMO antenna with different positions and orientations of metamaterial unit cells[61].

Fig. 2.29. A prototype of the proposed MIMO antenna with a double-layer “mushroom” isolator [62].

Fig. 2.30. Configuration of the proposed MIMO antenna with metamaterial polarization- rotator wall [63].

In 2019, Liu et al. [64] designed an extremely compact two-element linear polarized

MIMO antenna. A unit cell in the form of a double-layer short wire is used as a metasurface. The researchers used the main features of the metasurface, such as the compact size and effective decoupling, to suppress the coupling between two nearby

Bowtie antennas that were otherwise strongly coupled. In 2019, Mark et al. [65] investigated a metamaterial-superstrate-inspired MIMO antenna for gain and isolation improvements. A resonant cavity effect was generated by the superstrate layer to improve the antenna gain. The superstrate was made of a hexagonal nested ring structure, placed above the antenna and configured to limit the near field coupling between radiating elements, thereby enhancing the isolation. Jafargholi et al. [66] in 2019 studied the capability of a magnetodielectric superstrate to minimize the surface wave propagation.

It was demonstrated that a superstrate consisting of a capacitively loaded loop (CLL) metamaterial can exhibit a high degree of surface wave attenuation. This feature helped

42 to suppress the mutual coupling caused by the surface wave in proposed MIMO antennas.

The fabricated prototype of this design is shown in Fig. 2.31.

Fig. 2.31. The fabricated prototype of the proposed MIMO antennas showing capacitively loaded loop metamaterial [66].

Finally, a summarized state-of-art of the metamaterial method, containing the important previous works, is given in Table 2.3, which details the centre frequency, isolation, electrical size, separation distance between the radiating elements, antenna peak gain, antenna efficiency and ECC. The MIMO antenna structure explained in [60] is simple whilst it has the lowest value of efficiency. Further, [66] shows maximum isolation of more than 40 dB and lowest ECC value. However, [65] has achieved a maximum gain of

9.49 dB compared to the other designs presented in this section.

Table 2.3. Summarized states of the art on the metamaterial method.

Centre Peak Antenna Isolation Separation Ref. Frequency Electrical Size Gain Efficiency ECC Remarks (dB) distance (GHz) (dBi) (%) Low isolation at

[57] 2.59 ≥11 0.86λo ×0.43λo ×0.007λo 2.4 GHz and big size Two substrate, there is some [58] 3.53 >15 0.59λo ×1.06λo × 0.028λo 0.011λo 2 disagreement between the

43 simulated and the measured radiation pattern Narrow

bandwidth, [59] 2.45 20 0.41λo × 0.19λo × 0.016λo 0.11λo 60 three absorbers cells low efficiency and

[60] 2.345 >15 0.62 λo × 0.35 λo ×0.009 λo 1.3 37 <0.2 no separation distance Expensive substrate and the isolation is better [62] 2.42 >20 0.96λo×0.96λo× 0.02λo 0.44λo 4 <0.1 than 15 dB without decoupling

4 % reduction in efficiency with

[63] 60.5 >20 0.5λo 88 <0.1 metamaterial wall, overall size not given Bowtie dipole antenna, narrow [64] 2.5 >20 1.25λo ×1.25λo ×0.008λo 0.27λo 6.1 90 <0.15 band, superstrate and big size Big size, [65] 5.875 >17 1.37λo ×1.17λo ×0.046λo 0.0587λo 9.49 >84 <0.15 superstrate and high gain Big size, less than 15 dB before applying [66] 3.325 >40 0.9λo ×1.62λo ×0.008λo 12.3λo 8.2 97 <0.01 metamaterial. wide separation distance

2.6.4 Parasitic Element Method

The parasitic element is one of the methods that can be used to reduce mutual coupling between the antennas in a MIMO system by inserting an appropriate shape in a specific location to minimize the coupling between the radiating elements of the MIMO antenna.

Min, Kim, and Moon [67] in 2005 designed a MIMO antenna with four patch elements to cover the WLAN frequency band 5 GHz (Fig. 2.32). They also investigated the effects of the length and the number of transmission lines between the elements. Mak, Rowell, and Murch [68] in 2008 designed two monopole antennas with a parasitic element between them for field cancellation. These antennas were designed to cover the WLAN frequency band at 2.4 GHz for a universal serial bus (USB) dongle. Fig. 2.33 shows the configuration of the proposed MIMO antenna with the parasitic element.

Fig. 2.32. A prototype of the four-element MIMO antenna with parasitic elements [67].

Fig. 2.33. Configuration of the proposed MIMO antennas for USB WLAN dongle [68].

In 2009, Lee, Chen, and Hsu [69] designed a dual PIFA antenna to work in the WLAN frequency band (2.4). A new method was used to reduce the coupling by placing a folded resonator above the antenna, as illustrated in Fig. 2.34. This folded resonator can be considered as a parasitic element. Kang and Wong [70] in 2009 designed a dual-band

MIMO antenna with meander line resonator to work as a parasitic element between the radiating antennas, as seen in Fig. 2.35. This MIMO antenna was designed to cover

WLAN frequency bands (2.4/5.2/5.8 GHz).

Fig. 2.34. Configuration of the proposed MIMO antenna with folded resonator above the antenna [69].

Fig. 2.35. Configuration of the proposed MIMO antenna with dual-band resonator as a parasitic element [70].

In 2012, Li et al. [71] designed dual-slot element antennas with a parasitic element to reduce the mutual coupling between the elements, as shown in Fig. 2.36. This MIMO antenna is designed for mobile terminal applications; therefore, the authors have analyzed the effect of reducing mutual coupling on the channel capacity of the system. In the same year, 2012, Ayatollahi, Rao and Wang [72] designed two quarter-wavelength monopole

46 slots etched on the ground plane and a meandered slot cut between them to work as a parasitic element, as shown in Fig. 2.37.

Fig. 2.36. Configuration of the proposed MIMO antenna with parasitic elements [71].

Fig. 2. 37. Fabrication of the proposed MIMO antenna with a parasitic element in the form of a meandered slot cut [72].

Soltani and Murch [73] in 2015 designed two canonical radiating elements with the inclusion of a parasitic element to reduce the coupling. These elements can be replicated together to form MIMO antennas with a variable number of elements. This antenna was designed to work in a frequency band of 2.6 GHz. In 2017, Ding et al. [74] designed an

47 array using four broadband printed directive elements with parasitic strips as elements for

MIMO antenna applications. This antenna was presented to work in the WLAN frequency band (2.4 GHz): Fig. 2.38 illustrates the fabricated antenna. In 2019, Ghimire et al. [75] presented a compact high-isolation UWB MIMO antenna. In this design, a circular parasitic element was introduced at the bottom side of the radiating patch to create a reverse coupling and help to enhance the isolation at the upper part of the intended band.

On the other hand, the ground plane has a small rectangular notch to widen the impedance bandwidth: the design geometry is illustrated in Fig. 2.39. Amin et al. [76] in 2019 designed a compact four-port MIMO antenna for the UWB spectrum. A customized parasitic structure was utilized on the back layer of the substrate to fulfil the required isolation level. The antenna isolation was further improved by introducing a dumbbell- shaped structure in the partial ground plane. A photo of the fabricated design is shown in

Fig. 2.40. In the same year, 2019, Tang et al. [77] presented a four-port UWB-MIMO antenna. The main features of this design such as the ultra-wide impedance bandwidth and high isolation were achieved by introducing a circular patch and two smaller parasitic

U-shaped patches as well as a common square-ring ground structure, double decoupling branches, multi slits, and truncated corners. The geometry of the proposed MIMO antenna is demonstrated in Fig. 2.41.

Fig. 2.38. Fabrication of the proposed directive element MIMO antenna using four broadband printed elements with parasitic sub-elements [74].

Fig. 2.39. The geometry of the proposed MIMO antenna with parasitic element; (a) Top view, (b) Rear view [75].

Fig. 2.40. The fabricated design of the proposed MIMO antenna with parasitic element;

(a) Top layer, (b) Rear layer [76].

Fig. 2.41. The geometry of the proposed MIMO antenna; (a) Top layer, (b) bottom layer

[77].

Finally, a summarized state-of-art of the parasitic element method, containing the important previous works, is given in Table 2.4, which details the centre frequency, isolation, electrical size, separation distance between the radiating elements, antenna peak gain, antenna efficiency and ECC. In [75], high efficiency, gain and isolation has been obtained while the antenna geometry is complex. Further, [72] shows single band operation and it also achieves the lowest ECC value of 0.001.

Table 2.4. Summarized states of the art on the parasitic element method.

Centre Peak Antenna Isolation Separation Ref. Frequency Electrical Size Gain Efficiency ECC Remarks (dB) distance (GHz) (dBi) (%) High isolation, no efficiency, [67] 5.25 >20 1.15λo ×1.15λo× 0.5λo 6.7 no calculated thickness Complex [68] 2.45 >15 0.16λo ×0.32λo ×0.13λo 0.0294λo 2 60 geometry and high isolation High isolation and efficiency and no picture [69] 2.5 >15 0.625λo ×0.42λo ×0.013λo 0.118λo 1.4 80 for the fabricated design

50 High isolation and no picture 2.4 ≥3 ≥77 [70] >15 0.4λo ×0.72λo ×0.006λo 0.12λo for the 5.2 ≥4.3 ≥76 fabricated design High isolation

[71] 1.92 >20 0.6λo ×0.38λo ×0.005λo 0.17λo 80 and no gain and ECC

Big size and there is a wide difference [72] 2.6 >15 0.83λo ×0.47λo ×0.013λo 0.11λo 77.5 0.001 between the simulated and measured BW High isolation [73] 2.6 >20 0.34λo ×0.29λo ×0.013λo 0.017λo 2.5 85 and efficiency parasitic structure only [74] 2.6 >14 0.68 λo ×0.68 λo ×0.0064 λo 0.19λo 5.5 90 <0.003 within the individual elements Big size and

[75] 10.3 >20 2.2 λo ×2.2 λo ×0.055 λo 13.5 87 <0.007 high gain and efficiency Complex structure on the [76] 6.85 >15 0.91λo ×0.98λo ×0.023λo 0.48λo 2 93 <0.2 bottom layer and high efficiency

Big size, high efficiency and

[77] 10.5 >20 1.4λo ×1.4λo ×0.056λo 4 70 <0.03 fabricated design is not included

2.6.5 Defected Ground Structure Method

Recently, a defected ground structure technique has been used widely in single and multiple antenna systems due to its benefits of widening the bandwidth, decreasing the coupling and miniaturizing the radiating elements. In fact, this technique is based on making a “defect” in the ground plane which typically adds inductance and capacitance inside the structure of the antenna. Consequently, the antenna can be modified to fulfil the requirements of interest by changing these additional inductances and capacitances.

Enhancing the isolation inside the MIMO antenna structure at a certain frequency band can be achieved by adding extra inductances and capacitances which are generated by the

51 defected area within the ground plane. The function of this approach can be classified as a band stop filter between the radiating elements [32].

In 2007, Chiu et al. [78] designed two closely packed PIFA MIMO antennas sharing the same ground plane. A defected ground structure technique in the form of a set of slits was used to reduce mutual coupling between the radiating elements. The isolation was improved by 15 dB after applying the DGS. This MIMO antenna functions in the 2.27-

2.35 GHz frequency band: its configuration is illustrated in Fig. 2.42. Xiong and He [79] in 2009 designed a MIMO antenna of four elements and DGS was applied in this design in the form of etched slits: a 10 dB coupling reduction was achieved after applying the technique. This MIMO antenna was designed to cover the 2.4 GHz WLAN band: Fig.

2.43 shows its schematic structure. Li et al. [80] in 2012 applied DGS between two symmetric monopoles for mutual coupling reduction purposes. Two defected areas in the form of bent slits were etched in the ground plane and two triangles were cut from the ground to improve the S-parameters of the antenna, as shown in Fig. 2.44. The isolation has been enhanced by approximately 12 dB.

Fig. 2.42. Configuration of the two PIFAs with DGS [78].

Fig. 2.43. The schematic structure of the four-element MIMO antenna with a series of slits etched in the ground plane: (a) Side view, (b) Top view, (c) Bottom view [79].

Fig. 2.44. The configuration of the proposed MIMO antenna with DGS in the form of bent slits and two triangular cut-outs [80].

In 2015, Luo et al. [81] designed a 2×2 Ultra-Wideband MIMO antenna. A T-shaped slot was etched in the ground plane to work as DGS and hence reduce the mutual coupling between the radiating elements for the frequencies from 4 GHz upwards. The antenna achieved isolation of more than 18 dB over the operating frequency band (3.1-10.6 GHz):

Fig. 2.45 illustrates the schematic structure of the proposed antenna. Jilani and Alomainy

53 [82] in 2018 designed a millimetre-wave T-shaped MIMO antenna for fifth-generation

(5G) wireless applications. DGS in the form of split rings were etched in the ground plane to widen the antenna bandwidth. The practical results of this antenna showed a high bandwidth from 25 to 37.5 GHz with good isolation (Fig. 2.46).

Fig. 2.45. The schematic structure of the proposed UWB MIMO antenna with DGS in the form of a T-shaped slot [81].

Fig. 2.46. The configuration of the proposed mmWave T- shaped MIMO antenna: (a)

Top view of the simulated prototype, (b) Bottom view of the simulated prototype, (c) Top view of the fabricated prototype, (d) Bottom view of the fabricated prototype [82].

54 In 2019, Hasan et al. [83] presented 2×2 and 4×4 MIMO antennas using the DGS technique. In this design, the upper corner frequency of the antenna was extended to 11.5

GHz from 8 GHz by optimizing the position of two simple slits and a notch on the ground plane of the proposed antenna. Moreover, the single radiating element was loaded with a

U-shaped stub to improve the impedance matching. The fabricated 2×2 and 4×4 MIMO antennas are shown in Fig. 2.47. Arun and Alsath [84] in 2019 studied a 2×2 dual- polarised microstrip patch MIMO antenna. A DGS in the form of cross-connected octagonal shape was integrated into the ground plane to minimize the coupling within the antenna structure. Furthermore, the overall size of the antenna was reduced by sharing an annular-shaped aperture between the vertical and horizontal ports. The structure of the proposed antenna is illustrated in Fig. 2.48.

Fig. 2.47. Fabricated prototypes of the MIMO antenna proposals [83]; (a) 2×2 MIMO antenna and (b) 4×4 MIMO antenna.

Fig. 2.48. The prototype of the proposed 2×2 dual-polarized microstrip patch MIMO antenna; (a) Top view (b) Bottom view [84].

55 Pasumarthi et al [85] in 2020 designed a two-port tri-band MIMO antenna for WiMAX,

Radar altimeter and X-band applications. The MIMO antenna performance in terms of isolation was evaluated with two structures (1) DGS (2) DGS with Vias. Further isolation enhancement was observed by applying DGS with Vias. The fabricated MIMO antenna is shown in Fig. 2.49. In 2020, Biswas et al. [86] implemented a dual feed orthogonally polarized square patch antenna array. In this model, a simple DGS in the form of a rectangular-headed cross was etched in the ground plane underneath the square patch in order to obtain high port isolation between two orthogonal ports. Furthermore, the dual orthogonal element array was adopted for MIMO antenna systems where two slotted microstrip resonators were added between the two array structures for further isolation improvement, as shown in Fig. 2.50.

Fig. 2.49. MIMO antenna prototype; (a) Top layer (b) Bottom layer [85].

Fig. 2.50. Configuration of the proposed MIMO antenna integrated with two slotted microstrip resonators [86].

56 Finally, a summarized state-of-art of the DGS method, containing the important previous works, is given in Table 2.5, which details the centre frequency, isolation, electrical size, separation distance between the radiating elements, antenna peak gain, antenna efficiency and ECC. Compared to other structures [82] has the simplest geometry with big size and separation distance. Additionally, the highest of 25 dB is achieved in [79]. Further, [78] has the highest efficiency with complex geometry.

Table 2.5. Summarized states of the art on the DGS method.

Centre Peak Antenna Isolation Separation Ref. Frequency Electrical Size Gain Efficiency ECC Remarks (dB) distance (GHz) (dBi) (%) Complex geometry, 0.33λo ×0.33λo compact size and [78] 2.31 20 0.116 λo 3.66 88 ×0.012 λo high isolation and efficiency High isolation and 0.64λo ×0.48λo× no efficiency and [79] 2.42 >25 2.3 <0.022 0.015λo separation distance recorded Big and Complex 1.16 λo ×0.59 λo 0.083λo at geometry and no [80] 4.475 >18 <0.01 ×0.02 λo 2.5 GHz gain and efficiency recorded 0.59λo ×0.5λo Average High isolation and [81] 6.85 >18 45 <0.004 ×0.018λo gain 3.8 Low efficiency Big size and 5.2λo ×1.25λo separation distance, [82] 31.3 >20 1.3λo >4.5 >80 <0.06 ×0.08λo high isolation and expensive substrate Big size, No gain

and efficiency for

1.95 λo ×0.97 λo 2×2 and 4×4 [83] 7.34 >15 0.37 λo <0.015 ×0.038 λo MIMO, Only provided for single elements Compact size and

0.21λo ×0.21λo no efficiency and [84] 1.91 >15 3.63 0.004 ×0.01λo separation distance reported 2.9 No efficiency and 0.6 λo ×0.29 λo [85] 4.3 >15 0.087λo 0.0001 gain reported 7.7 ×0.015 λo Most of the

1.23λo ×0.55λo influencing [86] 2.875 >25 ×0.015λo parameters are not included

57 2.7 Comparison of the Different Mutual Coupling Reduction

Methods

A summarized comparison of the coupling reduction methods and isolation enhancement approaches that have been presented in this chapter in terms of their function, advantages and disadvantages is given in Table 2.6.

Table 2.6. Comparison of mutual coupling reduction methods.

Name of the Ref. Function Advantages Disadvantages method

Placing the radiating Antenna placement • Simple. • Needs extra space [33-46] elements at different and orientation • Good isolation. • Polarization mismatch. positions or orientations

Takes the current at a

specific location on the Simple structure. • • It is difficult to select radiating element and the connection points [47-56] Neutralization line • Good impedance matching. inverts its phase to cancel Narrow bandwidth. Good diversity gain. • the current from the •

adjacent radiating element.

• Needs to specify the unit

Produces a stop band at a • Reduces antenna size. cell features. certain frequency which • Improves channel capacity [57-66] Metamaterial • Leads to reduced leads to reduced coupling antenna efficiency. at that band. • Narrow bandwidth.

• Shift in the operating Creates an opposite • Simple structure. frequency. [67-77] Parasitic elements coupling field that reduces • Good diversity gain. • May lead to reduced the original one • Produces less ohmic loss. antenna efficiency.

Defected ground The defected area in the • Reduces antenna gain. [78-86] • Improved antenna BW. structure ground structure works as • Back-lobe radiation.

58 a band stop filter between • Produces multi-band or the radiating elements. wideband benefits.

• Reduces size.

2.8 MIMO Antennas by Band Classifications

Over the last decade, several significant and novel approaches to improve the isolation between MIMO antenna elements have been reported, as explained previously. These techniques and solutions were applied to several antenna designs with different band classifications such as narrow band, wideband and dual/triple band. These classifications are discussed in the following sub-sections.

2.8.1 Narrow Band MIMO Antennas with Reduced Mutual

Coupling

A remarkable amount of effort has been invested to design narrow band MIMO antennas.

Such antennas comprise several standards, starting from the existing generations up to the new and expected services. For example, in [87] MIMO antenna design for operation in the LTE bands was proposed; a GSM 800 MHz MIMO antenna was studied and investigated in [88, 89]; a UMTS MIMO antenna was studied in [71] and another MIMO antenna covering the lower and upper WLAN bands, namely, 2400 MHz and 5200 MHz, was presented in [50, 79, 90-101].

Moreover, several approaches to reduce the mutual coupling were applied to those MIMO antennas. A method called magneto-dielectric substrates was applied to an antenna in

[87], which hugely minimizes the mutual coupling, and the idea of introducing some

59 parasitic elements exactly between the two antennas was proposed in [71, 88]. The aid of using resonators was also seen as an effective avenue to improve the isolation between the antenna elements as in [89, 97, 98]; a novel approach of reducing the coupling was seen through the implementation of a neutralization line [50] and generating a form of slits over the PCB could also be a way to stop the current flow from one element to another [79]. Introducing lumped elements [90, 91, 99], a ground plane with substantial slots [54], an orthogonal arrangement of the antenna elements [87, 93, 96], a reconfigurable mechanism by using a varactor diode [94], a matching network in the ground plane [100] and a hybrid electric and magnetic approach in the form of the split ring [101] have also been reported, with good results. A few influencing parameters of some of these methods are listed in Table 2.7. It can be seen that these methods have accomplished good mutual coupling reduction, which ranges from 10 dB to 40 dB for reasonable inter-element distances from 0.03λo to 0.17 λo.

Table 2.7. Previous work on narrow-band MIMO antennas.

Ref. Separation Isolation/Return Size (mm3) BW(GHz) No. distance loss (dB) [88] 100×50×4.2 0.746-0.823 0.15λo 15/6

[89] 50×40×1.6 0.76-0.886 0.03λo 12/10

[71] 95×60×0.8 1.92-2.17 0.17λo 20/10

[90] 50×50×6 2.3-2.4 0.09λo 25/10

[50] 65×30×0.8 2.4-2.5 0.112λo 15/10

[79] 75×50×2.6 2.4-2.5 0.08λo 25/10

[91] 67.5×22×0.8 2.4-2.5 0.04λo 20/10

[92] 34.6×34.6×1 5.2-5.3 0.1λo 20/10

[93] 32×27.8×0.8 5.75-5.85 0.031λo 40/10

[94] 38×15×3.2 5.75-5.83 0.14λo 20/10

[95] 50×40×0.8 1.4-1.65 0.05λo 10/10

60 2.8.2 Wide Band MIMO Antennas with Reduced Mutual

Coupling

The key challenge here is to compromise between the operating frequency bandwidth, isolation level as well the limited space between the MIMO antenna elements. In other words, it is not desirable to achieve one of the above features at the cost of another one.

For example, it is not practical to come up with a wideband MIMO design that also achieves significantly high isolation across the band. In addition, having a bulky MIMO antenna design with a large distance between the elements will help in reducing the coupling; however, it is not practical to be incorporated within today’s smart devices. To meet this challenge, some methods to enable wideband operation, while the isolation is being improved, were reported in [102–110]. In [102], isolation as high as −11 dB at 0.25

λo spacing for a broadband range from 1.65 to 2.5 GHz has been achieved. Improved mutual coupling is observed within the very limited spacing distance by an orthogonal arrangement of the antenna elements [103], introducing inverted L-parasitic monopoles

[104], using the approach of neutralization lines between the antenna elements [105], using vertical branches between the radiators [106], using a coupled feed network [107], introducing inverted/non-inverted L and T ground branches [108], investigating pattern diversity [109] and applying the pattern diversity approach [110]. Table 2.8 summarizes the above-mentioned MIMO antenna designs.

61 Table 2.8. Previous work on wideband MIMO antennas.

Ref. Separation Isolation/Return Size (mm3) BW(GHz) No. distance loss (dB) [102] 67×50×0.254 1.65-2.5 0.25λo 11/10

[103] 86×86×1.52 1.63-2.05 0.1λo 20/10

[104] 95×60×0.8 1.85-2.17 0.063λo 14.8/10

[105] 90×40×0.8 2.4-4.2 0.19λo 17/10 [106] 25×39×0.8 2.68-12.8 >20/10

[107] 118×58×1.2 1.61-2.743 0.5 λo >12/6 [108] 110×65×0.8 1.7-2.9 >15/6 [109] 85×50×1.6 2-9.5 >20/10 [110] 40×40×1.6 3.1-10.6 ≥20/10

2.8.3 Multi-Band MIMO Antennas with Reduced Mutual

Coupling

Several dual- and triple-band MIMO antennas that demonstrate improved mutual coupling techniques have been proposed in [111–129]. It is seen that an improved isolation may be obtained by employing techniques within only a single band or by combining other bands. For instance, in [111] an approach of etching two inverted L- shaped lines together with a T-shaped branch over the ground plane shows reduced mutual coupling across the UMTS and WLAN 2.4 GHz bands. To simultaneously design a MIMO antenna capable of operating over the lower WLAN 2.4 GHz and WiMAX 3.5

GHz service, along with improved isolation, the authors in [112] achieved this by introducing a folded L-slot in the antenna surface and at the same time inserting a T- shaped slot over the antenna PCB.

62 A MIMO antenna operating over a wide frequency range of WLAN spectrum from

WLAN 2400 MHz to 5800 MHz, while achieving low mutual coupling, was proposed in

[113]: attenuation of mutual coupling was improved by carefully modifying the lengths of parasitic elements. A promising broadband MIMO antenna was investigated in [114].

It was suggested that a U-shaped slot was placed over the defected ground plane to perform as an effective decoupling method. On the other hand, the antenna design in

[115] utilizes three-dimensionally folded monopole elements in order to accomplish triple resonant modes together with an improved mutual coupling over the three allocated bands. The effect of shorting strips is presented in [116]: these strips help to create current loops which in turn control the directly coupled currents. The authors in [117] used different slot structures to improve the isolation and achieve multi-band operation. A defected ground structure in the form of a fork-shaped slot was introduced in [118, 119].

Antenna structure and differential arrangement decoupling approaches were used as isolation improvers in [120]. Adding neutralization lines between the antenna elements to improve the isolation has been widely reported [121-125]. Using two methods that integrates DGS with parasitic elements is reported in [126]. Furthermore, the authors in

[127] inserted a decoupling network between the two antennas without increasing the size of the structure: a high isolation, reaching to 30 dB was obtained. A high port-to-port isolation of two closely spaced inverted-F antennas was achieved by inserting a parasitic lossless radiating element between the two antennas with optimized position, dimensions, and shape [128]. The authors in [129] used a decoupling approach based on using arrays of capacitively loaded loops (CLLs) on the top layer of the PCB, with complementary

CLLs on the bottom layer of the board. The details of some significant dual- and triple- band MIMO antennas are given in Table 2.9.

63 Table 2.9. Previous work on multi-band antennas.

Ref. Separation Isolation/Return Size (mm3) BW(GHz) No. distance loss (dB) 1.92-2.17 13, 18/ [111] 95×60×0.8 2.4-2.5 0.038λo 10, 10 2.4-2.5 19.2, 22.8/ [112] 100×40×6 3.4-3.6 0.0088λo 10, 10 2.4-2.5 15, 22/ [113] 150×100×0.8 5.15-5.825 0.06λo 10, 6 1.5-2.8 15, 17 [114] 63×x50×0.8 4.7-8.5 0.13λo 10, 10 0.868-0.915 10, 10, 10/ [115] 105×55×7.5 1.7-2.3 0.05λo 2.4-3.5 6, 10, 10 2.39-2.68 25, 30/ [116] 100×60×0.8 4.85-6 0.0432λo 6, 6 2.35-2.5 33, 20/ [117] 17.3×22×7.7 5.19-5.4 0.0164λo 10, 10 2.3-2.5 22, 22/ [118] 100×50×0.8 5.75-5.85 0.25λo 10, 10 2.4-2.5 20, 20/ [119] 69×34×4.2 5.15-5.85 0.024λo 10, 10 2.45-2.55 3.4-3.6 35, 35, 40, 40/ [120] 60×40×NG 0.024λ 5-5.5 o 10,10, 10, 10 6.8-7.5 2.4-2.7 10, 15/ [121] 50×40×1.6 5.6-6.1 0.081λo 10, 10 2.39-2.53 19, 22/ [122] 50×26×0.8 4.57-6.09 0.098λo 10, 10 2.2-2.6 22, 18, 30/ [123] 30×50×0.8 3.4-4.2 0.096λo 5.7-6.1 10, 10, 10 2.24-2.45 14, 18, 16/ [124] 49×48×0.8 3.3-4 0.031λ o 10, 10, 10 5.6-5.75 2.17-2.61 15, 15/ [125] 38×36×1.6 5.53-5.87 0.0239λo 10, 10

2.9 Hybrid Method for Mutual Coupling Reduction

Most of the works discussed previously deal with one single method to enhance the performance of MIMO antenna systems in terms of isolation, bandwidth and correlation

64 coefficients. In many cases, using one single method does not do enough to enhance all of the influencing parameters that determine the antenna performance. Therefore, another important method that integrates two or more decoupling approaches within one MIMO antenna design has attracted researchers in this field. This method can easily help to design a MIMO antenna with high isolation and compact size. Many examples of using this hybrid method to enhance the isolation have been listed in the literature.

In 2012, Yao et al. [130] designed a broadband MIMO antenna for multimode satellite navigation. The designed antenna consists of circular polarized two-element PIFA. A hybrid technique that combines antenna orientation with cross stub approaches was used to improve the isolation between the two ports, as shown in Fig 2.51. Chacko, Augustin and Denidni [131] in 2014 presented a dual planar patch MIMO antenna with a combined decoupling approach (combining stub and antenna orientation methods) for UWB applications. The two radiating elements are placed perpendicularly to each other to minimize the coupling between the input ports while two long ground stubs are added to further improve the isolation, as shown in Fig. 2.52. Tripathi, Mohan and Yadav [132] in

2015 applied a hybrid technique that includes antenna orientation and stubs into a compact octagonal shaped fractal UWB MIMO antenna: Kock fractal geometry was applied to obtain miniaturization in size and wideband operation. Placing the two radiating elements in orthogonal geometry improved the antenna isolation while the grounded stubs were added to the structure to obtain further reduction in the coupling:

Fig. 2.53 illustrates the fabricated prototype of the antenna. In 2016, Singh, Shalini and

Meshram [133] presented a dual-band compact printed MIMO antenna for USB dongle application. An integrated method that combines neutralization line with defected ground

65 structure was used to improve the isolation to better than 14 dB at the two frequency bands. Fig. 2.54 illustrates the schematic structure of the proposed antenna and positions of NL and DGS.

Fig. 2.51. Fabricated prototype of the proposed MIMO antenna with cross stubs [130].

Fig. 2.52. Schematic structure of the proposed dual-element MIMO antenna with ground stubs [131].

Fig. 2.53. Fabricated prototype of the proposed compact octagonal shaped fractal MIMO antenna with ground stubs [132].

Fig. 2.54. Schematic structure of the proposed dual-band MIMO antenna with NL and

DGS [133].

In 2017, Cheung et al. [134] designed a planar dual-band MIMO antenna consisting of two PIFA antennas. A defected ground structure approach was combined with two meander lines inside the defected area for dual-band operation and mutual coupling reduction, as illustrated in Fig. 2.55.

Fig. 2.55. Configuration of the proposed MIMO antenna; (a) MIMO antenna, (b) PIFA element, and (c) DGS and two meander lines [134].

67 Thummaluru and Chaudhary [135] in 2017 integrated the DGS decoupling approach in the form of a rectangular cut between the radiating elements with a mu-negative metamaterial filter decoupling approach. The two combined approaches were used in the basic two port monopole antenna and then in a two port circularly polarized MIMO antenna, as illustrated in Fig. 2.56. Good isolation has been achieved after applying the two decoupling approaches. Mondal, Dhara and Harish [136] in 2018 designed a compact wideband MIMO antenna. In this design, the MIMO antenna comprised a wideband NL connected in between two symmetric slotted D-shaped monopoles and a DGS which was cut into the top of the ground plane as shown in Fig. 2.57. The neutralization line method was used as a decoupling technique while the defected ground structure acted as an impedance matching element for both ports.

Fig. 2.56. Fabricated prototypes of the proposed MIMO antennas with DGS and mu- negative metamaterial; (a) Basic two port monopole antenna, (b) Circularly polarised

MIMO antenna [135].

Fig. 2.57. Schematic structure of the proposed MIMO antenna with wideband NL and

DGS [136].

In 2019, Zhang et al. [137] presented a 2×2 MIMO antenna of planar monopoles for UWB applications. A combination method that combines three Archimedes spiral electromagnetic band gap (As-EBG) structures for band rejection purposes alongside a

T-shaped strip and triangular cut on the ground plane for mutual coupling reduction is shown in Fig 2.58. Luo et al. [138] in 2019 modelled a MIMO antenna array operating at

5.7 GHz for WLAN applications. In this module, a suspended meta-surface consisting of periodic metamaterial cells was applied on the top of the MIMO antenna array to realize a gain enhancement; on the other hand, a neutralization line decoupling structure was added on the feeding lines to improve the isolation, as shown in Fig. 2.59. Finally, it can be concluded from this survey that most of the MIMO designs had dealt with single band of operation while using an integrated decoupling method is limited in multi-band MIMO antennas for portable applications. Therefore, more work should be attempted to show its benefits on multi-band MIMO antennas MIMO antennas designs.

Fig. 2.58. Geometry of the proposed MIMO antenna with As-EBG, T-shaped strip and

DGS [137].

Fig. 2.59. Suspended meta-surface structure consisting of periodic metamaterial cells with neutralization line decoupling structure added on the feeding lines; (a) Top view, (b)

Rear view [138].

A summarized state-of-art of hybrid methods containing the important previous works is given in Table 2.5, which details the centre frequency, isolation, electrical size, separation distance between the radiating elements, antenna peak gain, antenna efficiency and ECC.

70 Table 2.10. Summarized states of the art on the hybrid method element method.

Centre Separation Peak Antenna Isolation Electrical Size Type of hybrid Ref. Frequency distance Gain Efficiency ECC Remarks (dB) method (GHz) (dBi) (%) Antenna No efficiency

[130] 1.375 >14 0.43λo × 0.43λo × 0.007λo <2.6 <0.02 orientation and and separation stub distance Big size, high Antenna isolation and no [131] 6.85 >25 0.93λo × 0.52λo × 0.017λo 0.068λo <4 orientation and efficiency and stub ECC 4 2.9 Antenna High isolation, [132] 7 >17 0.6λo × 0.6λo × 0.021λo 0.014λo 3.9 <0.01 orientation and good gain and no 10 4.5 stub efficiency 2.425 >15 0.18 65 0.16 Compact size [133] 3.55 >14 0.2λo × 0.48λo × 0.013λo 0.094λo 3.6 85 0.015 DGS and NL and low gain at 5.6 >15 2.74 84 0.001 the first band High isolation 2.17 60 DGS and two [134] >15 0.44λo × 0.47λo × 0.005λo 0.217λo and no peak gain 5.26 83 meander lines and ECC Simple structure,

DGS and high isolation [135] 2.675 >15 0.4λo × 0.4λo × 0.014λo 0.22λo <0.13 metamaterial and no efficiency and peak gain Simple structure,

<0.00 high isolation, [136] 11.9 >22 0.595λo × 0.87λo × 0.02λo 0.047λo 73 DGS and NL 2 low ECC and no peak gain EBG, T-shaped High isolation, [137] 6.85 >20 0.548λo × 0.77λo × 0.018λo 0.219λo 2.5 <0.03 strip and DGS good gain and no efficiency Big size, high Metamaterial and [138] 5.7 >30 1.5λo × 1.045λo × 0.66λo 0.095λo 8.3 68 isolation and NL gain

2.10 Identification of Gaps in Knowledge (Problem Statement)

Different isolation enhancement methods [33-138] have been reported to enhance the

performance of MIMO antenna systems in terms of correlation, isolation, and antenna

gain. Enormous efforts for mutual coupling suppression have been made. These methods

include antenna placement and orientation [33-46], use of neutralization line [47-56],

metamaterial resonators [57-66], parasitic elements [67-77], defected ground structure

[78-86] and hybrid method [130-138]. The analysis of these different methods has been

discussed in detail. It has been observed from this comprehensive survey that many

decoupling methods either have complex structures, which in turn lead to many

71 difficulties in their implementation stage, or still require to be improved without affecting the performance of the other adjusting modifications to fulfil the demands of the portable devices. Consequently, more effort in this research area is crucial to design new multiple antennas with a viable decoupling approach or combination of approaches that produce attractive characteristics for various modern wireless applications, for instance: high isolation, low correlation coefficient and simple and compact structure.

Furthermore, it has been found that the MIMO antenna technology has the following impeding aspects for which it is desirable that further research be undertaken:

1. A variety of MIMO antennas have unacceptably large sizes due to improper

utilization of substrate space.

2. Using an integrated decoupling method is limited in multi-band MIMO antennas

for portable applications.

3. Many designs do not discuss all the influencing parameters in detail.

2.11 Fundamental Parameters of MIMO Antenna Systems

Generally, the performance of a single antenna or antenna array is measured by analysing its S-parameters and gain patterns. These parameters are not sufficient to evaluate the performance of MIMO antenna systems. There are several reasons for this [6].

1. It is not convenient to analyse the S-parameters while the number of antennas

increases due to the increasing numbers of S-parameter curves. For example, a

system with two antennas has four S-parameters, but nine S-parameters for 3

antennas. In other words, there are N2 parameters for N antennas.

72 2. A MIMO antenna system does not depend only on the performance of its

elements, the mutual coupling between the elements should be considered.

3. It is necessary to take into consideration the effect of the channel where the

antenna system works.

Therefore, the following are the performance metrics used to evaluate MIMO antenna systems. [6]:

1. Total Active Reflection Coefficient.

2. Isolation.

3. Correlation Coefficient.

4. Diversity Gain.

5. System Channel Capacity.

6. Mean Effective Gain (MEG).

Each of these metrics will be discussed in detail.

2.11.1 Total Active Reflection Coefficient

Total Active Reflection Coefficient (TARC) indicates the importance of the stability in the resonance frequency even when the phase difference between antenna elements is changed. This factor can be calculated by taking the square root of the ratio between the available power minus the radiated power to the total available power as explained in equation (2.6) [139].

푡 available power ― radiated power 훤푎 = (2.6) available power

The range of TARC value is between 0 and 1. It can be considered that all the available power is radiated if the value of TARC is 0. The effective operating bandwidth of a

MIMO antenna can be extracted from TARC curves. One of the effective methods that is used to find the TARC from the measured S-parameters is explained in equations (2.7) to

(2.9).

푁 2 ∑ | | 푖 = 1 푏푖 푡 훤푎 = (2.7) 푁 2 ∑ | | 푖 = 1 푎푖

Where

b=Sa (2.8)

Where vector a represents the applied excitation to the antenna. For a two-port network, the TARC value can be calculated by the following expression [50].

푗휃 2 푗휃 2 ((|푆11 + 푆12푒 | ) + (|푆21 + 푆22푒 | )) 훤푎푡 = (2.9) 2

Where θ represents the phase of the input feeding.

74 2.11.2 Isolation

Isolation is defined as the degree of coupling between the radiating elements inside the

MIMO system structure [6]. This value does not depend on the coupling through the radiation pattern. Isolation is measured through the S-parameters of the system. The transmission coefficient (Sxy) between the two radiators’ feeding ports (radiator x and radiator y) measures this quantity. The isolation can be compromised by several factors.

The radiating elements can be coupled with each other through electric and/or magnetic fields within the antenna structure. Ground plane currents can also be a major factor in coupling the radiating elements as the ground plane size is small in compact antenna systems. There are a large number of techniques available in the literature that address this issue. Some of them use defects in the ground plane, parasitic elements in the antenna structure, metamaterial resonators, lumped components, frequency selective materials, neutralization line, magnetic wall between the radiators and many others.

2.11.3 Correlation Coefficient

Generally, wireless communication channels should be isolated from each other to prevent interference between them. For this reason, the correlation coefficient can be considered as a significant factor to measure the degree of isolation between the channels in MIMO systems. This factor depends on the radiation pattern of the MIMO antenna system. The envelope correlation coefficient (ECC) or (ρe) is the square of the correlation coefficient and can be calculated by the following expression [140], which assumes the isotropic scattering case:

75 2 |∬ [퐴 (휃,∅) ∙ 퐴 (휃,∅) ]푑Ω| 휌 = 4휋 1 2 (2.10) 푒 | |2 | |2 ∬ 퐴 (휃,∅) 푑Ω 퐴 (휃,∅) 푑Ω 4휋 1 2

Where the radiation pattern of the MIMO antenna is represented by the factor Ai (θ,∅).

This can be applied by terminating all the ports by matched loads and exciting the ith port only. This formula is valid for the isotropic environment. Furthermore, equation (2.10) is difficult and needs a 3D measurement for the radiation pattern as well as a numerical analysis to evaluate the value of ρe. The general expression has been simplified in order to calculate the correlation coefficient through the S-parameters and the radiation efficiency, as explained in the following formula [140].

∗ 2 2 |푆 푆 ∗ |휌 | = 휌 = 푖푖 푖푗 + 푆푗푖 푆 푗푗| (2.11) (푖푗) (푒푖푗) 2 2 2 2 | |(1 ― |푆 | ― |푆 | )(1 ― �| � | ― |푆 | )ɳ ɳ |(1 2)| 푖푖 푗푖 푗푗 푖푗 푟푎푑푖 푟푎푑푗

Where

ρij represents correlation coefficient.

ρeij represents envelope correlation coefficient.

Sij represents the S-parameter between the i and j elements.

th ηradi represents radiation efficiency for the i element.

Finally, high isolation does not necessarily mean that there is a low correlation. It is important to add here that the MIMO antenna system needs high isolation and low correlation coefficients to work perfectly.

76 2.11.4 System Channel Capacity

One of the important metrics in a MIMO antenna system is the channel capacity. This metric describes the difference between the performance of the single antenna system and the performance of the MIMO antenna system and how much enhancement has been achieved.

The measuring of the channel capacity is usually formed as a cumulative distribution function (CDF). Calculating the channel matrix is the first step to measure the channel capacity. The radiation patterns of the antenna system are used to determine the channel matrix. The formula that is used to determine the channel matrix is [71]:

1 1 2 2 퐻 = 휓푅 퐺휓푇 (2.12)

Where 휓R is the receive correlation coefficient matrix and 휓T is the transmit correlation coefficient matrix, while G is a random matrix containing complex Gaussian random numbers representing the randomness of the channel [6].

The entries for the correlation matrix can be evaluated as [141]:

휇푖,푗 푖,푗 휓푅 = (2.13) 휇푖,푖휇푗,푗

∗ ∗ 퐸 [퐴푖(Ω)ℎ(Ω)][퐴 (Ω)ℎ (Ω)] 푑Ω (2.14) 휇푖,푗 = ∫ { 푗 }

th The field pattern of the i elements is represented by Ai (Ω) and the incoming waves are represented by h (Ω) while Ω is the solid angle. There are reasonable assumptions for the environment of the mobile wireless systems as defined in [141]. These assumptions can be described as the fading envelope being Rayleigh distributed while the incoming wave should arrive in the horizontal plane only. On the other hand, the incoming wave’s

77 orthogonal polarizations should be uncorrelated while the individual polarizations should be spatially uncorrelated. Finally, the time-averaged power density per steradian should be constant [71]. From these assumptions:

퐴(Ω) = 퐴 (Ω)휃 + 퐴 (Ω)∅ 휃 휑 (2.15)

ℎ(Ω) = ℎ휃(Ω)휃 + ℎ휑(Ω)∅ (2.16)

Where 휃 and ∅ are the unit vectors along θ and ∅ axes.

Substituting equations (2.15) and (2.16) into equation (2.14) will lead to:

∗ ∗ ∗ ∗ 휇 = 퐸 {[퐴푖휃(Ω)ℎ휃(Ω) + 퐴푖∅(Ω)ℎ∅(Ω)] × [퐴 (Ω)ℎ (Ω) + 퐴 (Ω)ℎ (Ω)]}푑Ω 푖,푗 ∫ 푗휃 휃 푗∅ ∅ (2.17)

It can be expected that the orthogonal polarizations’ product is equal to zero because the incoming wave’s orthogonal polarizations are uncorrelated. Equation (2.17) can be simplified to:

∗ 2 ∗ 2 휇푖,푗 = {퐸 [퐴푖휃(Ω)퐴 (Ω)|ℎ휃(Ω)| ] + 퐸[퐴푖∅(Ω)퐴 (Ω)|ℎ∅(Ω)| ]}푑Ω ∫ 푗휃 푗∅ (2.18)

Moreover, due to the performance expectations being in terms of incoming waves, equation (2.18) can be simplified to:

∗ 2 ∗ 2 휇 = {퐴푖휃(Ω)퐴 (Ω)퐸 [|ℎ휃(Ω)| ] + 퐴푖∅(Ω)퐴 (Ω)퐸[|ℎ∅(Ω)| ]} 푑Ω 푖,푗 ∫ 푗휃 푗∅ (2.19)

78 2 2 Equation (2.19) can be further simplified due to the fact that 퐸 [|ℎ휃(Ω)| ] and 퐸[|ℎ∅(Ω)| ] are the time-averaged power density of the incoming waves of components 휃 and 휑

2 2 respectively. On the other hand, the ratio between 푡ℎ푒푚, 퐸 [|ℎ휃(Ω)| ]/퐸[|ℎ∅(Ω)| ] is defined as a cross-polarization power ratio (XPD) that represents the distribution of the incoming waves [142]. Moreover, the incoming wave arrives in the horizontal plane only and the time averaged power density per steradian is constant. Equation (2.19) can thus be written as: 휋 ∗ ∗ [XPD 퐴푖휃 , ∅ 퐴 , ∅ + 퐴 , ∅ 퐴 , ∅ ]푑∅ 휇푖,푗 = 푐 2휋 ( 휋 ) ( 휋 ) 푖∅( ) ( 휋 )

∫ 2 푗휃 2 2 푗∅ 2 (2.20) 0

Where c is a constant linked with the time-averaged power density. It is important to recognize that normalization is imposed in equation (2.14). Thus, c can be set to unity in equation (2.20).

After all these calculations, H can be easily found. The next step is to evaluate the capacity based on the following formula:

휌 푇 퐶 = log2 det 퐼푅 + 퐻퐻 (2.21) [ ( 푁푇 )]

Where IR is an identity matrix, NR represents the receive antennas and NT the transmit antennas, while HT is the conjugate transpose of the H matrix. If H is an identity matrix, then it is clear to say that the capacity is the number of antennas (rank of H) times the capacity of one antenna.

79 Theoretically, by increasing the number of antennas of the MIMO system, the channel capacity will improve [36]. However, the presence of uncorrelated Rayleigh fading will induce loss of channel capacity. This loss can be obtained from the correlation matrix. In the case of 2 x 2 MIMO systems, the simplified channel capacity loss of high SNR can be evaluated by using the following equation [143]:

∁ = ―log det (휓 ) 푙표푠푠 2 R (2.22) where 휓R is the receiving antenna correlation matrix assuming complete decorrelation at the transmit end:

휌11 휌12 휓 = 푅 [휌 21 휌22]

2 2 ∗ ∗ with 휌11 = (1 ― (|푆11| + |푆12| )) and 휌12 = ― (푆 11푆 12 + 푆 21푆 22)

2.11.5 Mean Effective Gain

Generally, testing an antenna gain inside an anechoic chamber is not a good measure of antenna performance because the antenna is not used in practical applications. The antenna will actually be used in a certain environment for a specific application.

Therefore, to evaluate the antenna performance, the effect of the environment on antenna radiation characteristics should be studied. The only way to achieve this is by fabricating an antenna and testing it in known environment conditions with other antennas, having known characteristics. After that, the antenna performance should be recorded. This process must be repeated by tuning another antenna and testing it again to get the required design. It is clear that this procedure is time consuming and costly. This problem has a

80 solution that was proposed in [18]. In this proposal, the author suggests a probabilistic model for the environment. Hence the MEG can be calculated numerically by using the

3D radiation pattern with the suggested model within a mathematical formula. This numerical method helps to evaluate MEG by using the simulated or the measured gain patterns along with a model of the environment suitable for the application for which the antenna is being designed. MEG can be expressed mathematically by equations (2.18) and (2.19):

2휋 휋 XPD 1 MEG = ∫∫{ XPD + 1 퐺휃(휃,∅)푃휃(휃,∅) + XPR + 1 퐺∅(휃,∅)푃∅(휃,∅)} (2.23) 0 0

under the conditions:

2휋 휋 { } ∫∫ 퐺휃(휃,∅) + 퐺∅(휃,∅) sin 휃푑휃푑∅ = 4휋 0 0 (2.24) 2휋 휋 2휋 휋

∫∫푃휃(휃,∅)sin 휃푑휃푑∅ = ∫∫푃∅(휃,∅)sin 휃푑휃푑∅ = 1 0 0 0 0

Vertical mean incident power XPD = (2.25) Horizontal mean incident power

Where:

Gθ (θ,∅) and GØ (θ, ∅) are antenna gain components.

Pθ (θ,∅) and PØ (θ, ∅) represent the channel model.

81 2.11.6 Diversity Gain

Another important parameter that is used to evaluate the performance in MIMO antenna systems is the diversity gain. This metric can be used to measure the influence of the diversity on the wireless communication system performance. In other words, it describes the increase in signal to noise ratio when the diversity antenna system is applied [144].

Moreover, high diversity gain will lead to a low correlation coefficient and vice versa.

The diversity gain can be calculated mathematically as follows [144]:

훾푐 훾1 DG = (2.26) [ ― ] 훤푐 훤1 푝(훾푐 < 훾푠/훤)

Where

• 훾c is the instantaneous SNR for the diversity system and 훤푐 is mean SNR for the

diversity system.

• 훾1 is the instantaneous SNR for the single branch with maximum value and 훤1 is

the mean SNR for the single branch with maximum value

• 훾s/훤 is the reference level.

The factor 푝(훾푐 < 훾푠/훤) can be approximated by assuming uncorrelated signals with

Rayleigh distribution as follows:

훾푠 푀 훾푠 — 푃 훾 < = (1 ― 푒 훤 ) (2.27) ( 푐 훤 )

82 It is crucial to mention here that the diversity gain and correlation coefficient are related to each other. If the correlation coefficient is low, the diversity gain will be high. Equation

(2.27) is usually valid if |MEG1˗ MEG2| < 3 dB [111].

2.12 Conclusions

In this chapter, the main concept of MIMO technology has firstly been presented following by the theoretical background of MIMO technology and then a comprehensive review of MIMO antennas has been presented and many important aspects related to

MIMO antennas discussed. Discussion ranged from the history of MIMO antennas to their main advantages, and difficulties faced by researchers during their practical implementation. Also, this review focused on five different techniques that have been used in this thesis to minimize the mutual coupling in MIMO antennas. These techniques include antenna placement and orientation, metamaterial resonators, parasitic elements, neutralization line and defected ground structures. A comprehensive analysis of the sources to classify MIMO antennas according to their frequency bands (narrow bands, wideband and multi-band) has been presented in depth. The impact of combining two or more decoupling methods has been presented by reviewing the most relevant works in this field. It has been noticed that from this comprehensive survey that many decoupling approaches either have complex geometries, which in turn lead to many difficulties in their implementation stage, or still require to be enhanced without affecting the performance of the other adjusting modifications to achieve the demands of the portable devices.

83 Finally, It has been explained how the fundamental parameters of single antennas are not enough to evaluate the performance of the MIMO antenna systems. Therefore, different performance metrics related to MIMO antenna systems have been discussed to show their effect on MIMO antenna performance. The level of the isolation between the elements shows the extent to which the elements are isolated within the configuration of the antenna while the radiation patterns of the radiating elements depend on the value of the correlation coefficients to specify how much they are isolated in the environment. On the other hand, MEG describes the behaviour of the antenna system within a specific environment and TARC value indicates how much available power is radiated. Moreover, the channel capacity shows the improvement value in the data rate. The diversity gain has also been discussed due to its ability to identify how much the diversity affects the performance of the communication system.

References

[1] C. Oestges, M. Guillaud, and M. Debbah, “Multi-Polarized MIMO Communications:

Channel Model, Mutual Information and Array Optimization,” IEEE Wireless

Communications and Networking Conference, pp. 1057–1061, 2007.

[2] M. Sanchez-Fernandez, E. Rajo-Iglesias, O. Quevedo-Teruel, and M. L. Pablo-

Gonzalez, “Spectral Efficiency in MIMO Systems Using Space and Pattern

Diversities Under Compactness Constraints,” IEEE Transactions on Vehicular

Technology, vol. 57, no. 3, pp. 1637–1645, May 2008.

[3] X. Wang, Z. Feng, and K. M. Luk, “Pattern and Polarization Diversity Antenna with

High Isolation for Portable Wireless Devices,” IEEE Antennas and Wireless

Propagation Letters, vol. 8, pp. 209–211, 2009.

[4] I. Nadeem and D. Choi, “Study on Mutual Coupling Reduction Technique for MIMO

Antennas,” in IEEE Access, vol. 7, pp. 563-586, 2019.

[5] M. R. Sharawi, “Printed MIMO antenna engineering,” Norwood: Artech House,

2014.

[6] T. Brown, P. Kyritsi and E. D. Carvalho, “Practical Guide to MIMO Radio Channel

with MATLAB Examples,” John Wiley & Sons, 2012.

[7] C. E. Shannon, “Communication in the presence of noise,” Institute of Radio

Engineers, vol. 37, no. 1, pp. 10–21, Jan. 1947.

[8] R. Mathar and A. Schmeink, “Increasing signaling power not necessarily improves

channel capacity,” in International conference on signal processing and

communication systems (ICSPCS), 2010, pp. 1–5.

85 [9] V. Hinostroza and S. Salous, “Frequency Selective Fading on Wideband Mobile

Radio Channels,” in IEEE Vehicular Technology Conference, Sep. 2006, pp. 1–4.

[10] Foschini, G. J. and M. J. Gans, “On limits of wireless communications in a fading

environment when using multiple antennas,” Wireless Personal Communications,

vol. 6, no. 3, 1998, pp. 311–335.

[11] G. D. Golden, G. J. Foschini, R. A. Valenzuela, and P. W. Wolniansky, “Detection

algorithm and initial laboratory results using the V-BLAST space-time

communication architecture,” Electronics Letters, vol. 35, Jan. 1999, pp. 14–15.

[12] P. W. Wolniansky, G. J. Foschini, G. D. Golden and R. A. Valenzuela, “V-BLAST:

an architecture for realizing very high data rates over the rich-scattering wireless

channel,” 1998 URSI International Symposium on Signals, Systems, and Electronics.

Conference Proceedings (Cat. No.98EX167), Pisa, Italy, 1998, pp. 295-300, doi:

10.1109/ISSSE.1998.738086.

[13] G. J. Foschini, D. Chizhik, M. J. Gans, C. Papadias and R. A. Valenzuela, “Analysis

and performance of some basic space-time architectures,” in IEEE Journal on

Selected Areas in Communications, vol. 21, no. 3, pp. 303-320, April 2003, doi:

10.1109/JSAC.2003.809456.

[14] L. Zheng and D. N. C. Tse, “Diversity and multiplexing: a fundamental tradeoff in

multiple-antenna channels,” in IEEE Transactions on Information Theory, vol. 49,

no. 5, pp. 1073-1096, May 2003, doi: 10.1109/TIT.2003.810646.

[15] Walter da C. Freitas Jr., Francisco R. P. Cavalcanti, and Renato R. Lopes, “Hybrid

Transceiver Schemes for Spatial Multiplexing and Diversity in MIMO Systems,”

Journal of communication and information systems, vol. 20, no. 3, pp. 63-76, 2005.

86 [16] M. A. Jan, “Design and Fabrication of a MIMO Antenna System for 4G Wireless

Handsets,” M.Sc. thesis, King Fahd University of Petroleum and Minerals, Saudi

Arabia, June 2011.

[17] V. Plicanic, B. K. Lau, A. Derneryd, and Z. Ying, “Actual diversity performance of

a multiband diversity antenna with hand and head effects,” IEEE Transactions on

Antennas and Propagation, vol. 57, no. 5, pp. 1547-1556, May 2009.

[18] T. Taga, “Analysis for mean effective gain of mobile antennas in land mobile radio

environments,” in IEEE Transactions on Vehicular Technology, vol. 39, no. 2,

pp. 117-131, May 1990, doi: 10.1109/25.54228.

[19] A. R. Mehrotra, D. R. Jackson, J. T. Williams and S. A. Long, “An annular-ring

reduced surface wave microstrip antenna,” IEEE Antennas and Propagation Society

International Symposium. 1999 Digest. Held in conjunction with: USNC/URSI

National Radio Science Meeting (Cat. No.99CH37010), Orlando, FL, 1999, pp. 810-

813 vol.2, doi: 10.1109/APS.1999.789437.

[20] J. L. Allen and B. L. Diamond, “Mutual Coupling in Array Antennas”, Lincoln

Laboratory, M.I.T., Tech. Rep. Technical Rep. 424 (ESD-TR-66–443), 1966.

[21] K. Malkomes, “Mutual coupling between the elements of microstrip patch

antennas,” Electronics Letters, Vol.18, No.122, June 1982, PP. 520-522.

[22] E.Penard and J.P.Daniel, “Mutual coupling between the elements of microstrip

antennas,” Electronics Letters, Vol.18, No.4, July 1982, PP. 605-607.

[23] E. Van Lil and A. Van De Capelle, “Transmission line model for mutual coupling

between microstrip antennas,” in IEEE Transactions on Antennas and

Propagation, vol. 32, no. 8, pp. 816-821, Aug 1984.

[24] C. A. Balanis, “Antenna Theory: Analysis and Design,” New York: Wiley, 1997.

87 [25] R. R. Ramirez and F. De Flaviis, “A mutual coupling study of linear polarized

microstrip antennas for use in BLAST wireless communications architecture,” IEEE

Antennas and Propagation Society International Symposium. Transmitting

Waves of Progress to the Next Millennium. 2000 Digest. Held in conjunction

with: USNC/URSI National Radio Science Meeting, Salt Lake City, UT, 2000,

pp. 490-493 vol.2, doi: 10.1109/APS.2000.875177.

[26] T. Svantesson and A. Ranheim, “Mutual coupling effects on the capacity of

multielement antenna systems,” 2001 IEEE International Conference on Acoustics,

Speech, and Signal Processing. Proceedings (Cat. No.01CH37221), Salt Lake City,

UT, 2001, pp. 2485-2488 vol.4.

[27] J. W. Wallace and M. A. Jensen, “The capacity of MIMO wireless systems with

mutual coupling,” Proceedings IEEE 56th Vehicular Technology Conference, 2002,

vol.2, pp. 696-700.

[28] Y. Ge, K. P. Esselle, and Trevor S. Bird “E-Shaped Patch Antennas for High-Speed

Wireless Networks,” IEEE Transactions on Antennas and Propagation, vol.52,

no.12, pp. 3213-3219, 2004.

[29] H. Ndoumb`e Mbonjo Mbonjo, G. Wu, and V.Hanse, “Field- Theoretical

Investigations of the Influence of Mutual Coupling Effects on the Capacity of MIMO

Wireless Links,” Advances in Radio Science, 3, 2005, pp. 437–440.

[30] H. T. Hui, “Decoupling Methods for the Mutual Coupling Effect in Antenna Arrays:

A Review,” Recent Patents on Engineering, 2007, PP. 187-193.

[31] A. A. S. Ali, Sharlene Thiagarajah, “A Review on MIMO Antennas Employing

Diversity Techniques,” Proceedings of the International Conference on Electrical

88 Engineering and Informatics Institut Teknologi Bandung, Indonesia, June 17-19,

2007, pp. 388- 391.

[32] A. B. Numan, “Isolation enhancement techniques for printed MIMO antenna

systems,” M.Sc. thesis, King Fahd University of Petroleum and Minerals, Saudi

Arabia, April 2013.

[33] H. Carrasco, H. D. Hristov, R. Feick, and D. Cofré, “Mutual coupling between planar

inverted-F antennas,” Microwave and Optical Technology Letters, vol. 42, no. 3, pp.

224–227, August 15, 2004.

[34] A. I. Najam, Y. Duroc and S. Tedjini, “Multiple-Input Multiple-Output antennas for

ultra wideband communication,” INTECH Open Access Publisher, 2012.

[35] K.-L. Wong, J.-H. Chou, S.-W. Su, and C.-M. Su, “Isolation between GSM/DCS and

WLAN antennas in a PDA phone,” Microwave and Optical Technology Letters, vol.

45, no. 4, pp. 347–352, May 2005.

[36] S. H. Chae, S. Oh and S. Park, “Analysis of Mutual Coupling, Correlations, and

TARC in WiBro MIMO Array Antenna,” in IEEE Antennas and Wireless

Propagation Letters, vol. 6, pp. 122-125, 2007.

[37] S. Park and C. Jung, “Compact MIMO antenna with high isolation performance,”

Electronics Letters, vol. 46, no. 6, pp. 390–391, 2010.

[38] S. Zhang, P. Zetterberg and S. He, “Printed MIMO antenna system of four closely-

spaced elements with large bandwidth and high isolation,” in Electronics Letters,

vol. 46, no. 15, pp. 1052-1053, 22 July 2010.

[39] M. K. T. Al-Nuaimi and W. G. Whittow, “Performance investigation of a dual

element IFA array at 3 GHz for MIMO terminals,” in 2011 Loughborough Antennas

& Propagation Conference. IEEE, Nov. 2011, pp. 1–5.

89 [40] J. Xiong, M. Zhao, H. Li, Z. Ying and B. Wang, “Collocated Electric and Magnetic

Dipoles With Extremely Low Correlation as a Reference Antenna for Polarization

Diversity MIMO Applications,” in IEEE Antennas and Wireless Propagation

Letters, vol. 11, pp. 423-426, 2012.

[41] S.Y. Lin and I.H. Liu, “Small inverted-U loop antenna for MIMO applications,”

Progress In Electromagnetics Research, 34, pp.69-84, 2013.

[42] G. T. Jeong, S. Choi, K. H. Lee, andW. S. Kim, “Low profile dual wideband MIMO

antenna with low ECC for LTE and Wi-Fi applications,” International Journal of

Antennas and Propagation, Hindawi Publishing Corporation, vol. 2014, pp. 1-6,

2014.

[43] M. U. Khan, M. S. Sharawi, “A dual band microstrip annular slot based MIMO

antenna system,” Microwave & Optical Technology Letters, vol. 57, no. 2, pp. 360-

364, Feb. 2015.

[44] L. Malviya, R. K. Panigrahi, and M. V. Kartikeyan, “A 2×2 dual-band MIMO

antenna with polarization diversity for wireless applications,” Progress In

Electromagnetic Research C, vol. 61, pp. 91–103, Jan. 2016.

[45] R. Mathur and D. Santanu, “Compact 4-Port MIMO/Diversity Antenna with Low

Correlation for UWB Application,” Frequenz 2018.

[46] M. S. Aw, K. Ashwath, and T. Ali. “A compact two element MIMO antenna with

improved isolation for wireless applications,” Journal of Instrumentation 14, no. 06

(2019): P06014.

[47] A. Diallo, C. Luxey, P. Le Thuc, R. Staraj, and G. Kossiavas, “Study and Reduction

of the Mutual Coupling Between Two Mobile Phone PIFAs Operating in the

90 DCS1800 and UMTS Bands,” IEEE Transactions on Antennas and Propagation,

vol. 54, no. 11, pp. 3063–3074, Nov. 2006.

[48] S. Ranvier, C. Luxey, P. Suvikunnas, R. Staraj and P. Vainikainen, “Capacity

enhancement by increasing both mutual coupling and efficiency: a novel

approach,” 2007 IEEE Antennas and Propagation Society International

Symposium, Honolulu, HI, 2007, pp. 3632-3635.

[49] Z. Li, M. S. Han, X. Zhao and J. Choi, “MIMO antenna with isolation enhancement

for wireless USB dongle application at WLAN band,” 2010 Asia-Pacific Microwave

Conference, Yokohama, 2010, pp. 758-761.

[50] S.-W. Su, C.-T. Lee, and F.-S. Chang, “Printed MIMO-Antenna System Using

Neutralization-Line Technique for Wireless USB-Dongle Applications,” IEEE

Transactions on Antennas and Propagation, vol. 60, no. 2, pp. 456–463, Feb. 2012.

[51] E. Elkhazmi, C. H. See, N. A. Jan, R. A. Abd-Alhameed, N. Ali and N. J. McEwan,

“Design of a wideband printed MIMO monopole antenna using Neutralization lines

technique,” Asia-Pacific Microwave Conference, Sendai, Japan, 2014, pp. 983-

985.

[52] S. Zhang and G. F. Pedersen, “Mutual Coupling Reduction for UWB MIMO

Antennas with a Wideband Neutralization Line,” in IEEE Antennas and Wireless

Propagation Letters, vol. 15, pp. 166-169, 2016.

[53] H. Shin, W. S. Kibria and M. T. Islam, “Hexa band MIMO antenna with

neutralization line for LTE mobile device application,” Microwave and Optical

Technology Letters, vol. 58, no. 5 (2016): 1198-1204.

[54] J. Banerjee, R. Ghatak and A. Karmakar, “A compact planar UWB MIMO diversity

antenna with Hilbert fractal neutralization line for isolation improvement and dual

91 band notch characteristics,” 2018 Emerging Trends in Electronic Devices and

Computational Techniques (EDCT), Kolkata, 2018, pp. 1-6.

[55] C. Jui‐Hung, et al. “Dual‐band WLAN MIMO antenna with a decoupling element

for full‐metallic bottom cover tablet computer applications,” Microwave and Optical

Technology Letters, vol. 60, no. 5, pp. 1245-1251, 2018.

[56] RN Tiwari, P. Singh, BK. Kanaujia, “A compact UWB MIMO antenna with

neutralization line for WLAN/ISM/mobile applications,” International Journal of

RF and Microwave Computer‐Aided Engineering, vol. 29, no. 11, 2019.

[57] C. C. Hsu, K. H. Lin, H. L. Su, H. H. Lin and C. Y. Wu, “Design of MIMO antennas

with strong isolation for portable applications,” 2009 IEEE Antennas and

Propagation Society International Symposium, Charleston, SC, pp. 1-4, 2009.

[58] Y. Lee, H. Chung, J. Ha and J. Choi, “Design of a MIMO antenna with improved

isolation using meta-material,” 2011 International Workshop on Antenna

Technology (iWAT), Hong Kong, 2011, pp. 231-234.

[59] H. S. Lee and H. M. Lee, “Isolation improvement between loop antennas with

absorber cells,” 2011 IEEE International Symposium on Antennas and Propagation

(APSURSI), Spokane, WA, 2011, pp. 1735-1738.

[60] Youngki Lee, Deukhyeon Ga and J. Choi, “Design of a MIMO Antenna with

Improved Isolation Using MNG Metamaterial,” International Journal of Antennas

and Propagation, vol. 2012, Jul. 2012.

[61] D. A. Ketzaki and T. V. Yioultsis, “Metamaterial-Based Design of Planar Compact

MIMO Monopoles,” in IEEE Transactions on Antennas and Propagation, vol.

61, no. 5, pp. 2758-2766, May 2013.

92 [62] G. Zhai, Z. N. Chen and X. Qing, “Enhanced Isolation of a Closely Spaced Four-

Element MIMO Antenna System Using Metamaterial Mushroom,” in IEEE

Transactions on Antennas and Propagation, vol. 63, no. 8, pp. 3362-3370, Aug.

2015.

[63] M. Farahani, J. Pourahmadazar, M. Akbari, M. Nedil, A. R. Sebak and T. A. Denidni,

“Mutual Coupling Reduction in Millimeter-Wave MIMO Antenna Array Using a

Metamaterial Polarization-Rotator Wall,” in IEEE Antennas and Wireless

Propagation Letters, vol. 16, pp. 2324-2327, 2017.

[64] F. Liu, J. Guo, L. Zhao, X. Shen and Y. Yin, “A Meta-Surface Decoupling Method

for Two Linear Polarized Antenna Array in Sub-6 GHz Base Station Applications,”

in IEEE Access, vol. 7, pp. 2759-2768, 2019.

[65] R. Mark et al., “Isolation and Gain Enhancement Using Metamaterial Based

Superstrate for MIMO Applications,” Radioengineering 28, no. 4 (2019): 689.

[66] A. Jafargholi, A. Jafargholi and J. H. Choi, “Mutual Coupling Reduction in an Array

of Patch Antennas Using CLL Metamaterial Superstrate for MIMO Applications,”

in IEEE Transactions on Antennas and Propagation, vol. 67, no. 1, pp. 179-189, Jan.

2019.

[67] K.-S. Min, D.-J. Kim, and Y.-M. Moon, “Improved MIMO Antenna by Mutual

Coupling Suppression between Elements,” The European Conference on Wireless

Technology, 2005, Paris, 2005, pp. 125-128.

[68] A. C. K. Mak, C. R. Rowell and R. D. Murch, “Isolation Enhancement Between Two

Closely Packed Antennas,” in IEEE Transactions on Antennas and Propagation,

vol. 56, no. 11, pp. 3411-3419, Nov. 2008.

93 [69] C.-H. Lee, S.-Y. Chen, and P. Hsu, “Integrated Dual Planar Inverted-F Antenna with

Enhanced Isolation,” IEEE Antennas and Wireless Propagation Letters, vol. 8, pp.

963–965, 2009.

[70] T.-W. Kang and K.-L. Wong, “Isolation improvement of WLAN internal laptop

computer antennas using dual-band strip resonator,” in 2009 Asia Pacific Microwave

Conference. IEEE, pp. 2478–2481, Dec. 2009.

[71] Z. Li, Z. Du, M. Takahashi, K. Saito, and K. Ito, “Reducing Mutual Coupling of

MIMO Antennas with Parasitic Elements for Mobile Terminals,” in IEEE

Transactions on Antennas and Propagation, vol. 60, no. 2, pp. 473–481, Feb. 2012.

[72] M. Ayatollahi, Q. Rao and D. Wang, “A Compact, High Isolation and Wide

Bandwidth Antenna Array for Long Term Evolution Wireless Devices,” in IEEE

Transactions on Antennas and Propagation, vol. 60, no. 10, pp. 4960-4963, Oct.

2012.

[73] S. Soltani and R. D. Murch, “A Compact Planar Printed MIMO Antenna Design,”

in IEEE Transactions on Antennas and Propagation, vol. 63, no. 3, pp. 1140-

1149, March 2015.

[74] K. Ding, C. Gao, D. Qu and Q. Yin, “Compact Broadband MIMO Antenna with

Parasitic Strip,” in IEEE Antennas and Wireless Propagation Letters, vol. 16,

pp. 2349-2353, 2017.

[75] J. Ghimire, K.-W. Choi, and D.-Y. Choi, “Bandwidth enhancement and mutual

coupling reduction using a notch and a parasitic structure in a UWB-MIMO

antenna,” International Journal of Antennas and Propagation, Vol. 2019, No. 5, 1–

9, 2019.

94 [76] F. Amin, R. Saleem, T. Shabbir, S. Rehman, M. Bilal, and M. F. Shafique, “A

compact quad-element UWB-MIMO antenna system with parasitic decoupling

mechanism,” Applied Sciences, vol. 9, no. 11, p. 2371, Jun. 2019.

[77] Z. Tang, J. Zhan, X. Wu, Z. Xi, and S. Wu, “Simple ultra-wider-bandwidth MIMO

antenna integrated by double decoupling branches and square-ring ground

structure,” Microwave and Optical Technology Letters, vol. 62, no. 3, pp. 1259–

1266, Mar. 2019.

[78] C.-Y. Chiu, C.-H. Cheng, R. D. Murch, and C. R. Rowell, “Reduction of Mutual

Coupling between Closely-Packed Antenna Elements,” in IEEE Transactions on

Antennas and Propagation, vol. 55, no. 6, pp. 1732–1738, Jun. 2007.

[79] H. Li, J. Xiong, and S. He, “A Compact Planar MIMO Antenna System of Four

Elements with Similar Radiation Characteristics and Isolation Structure,” in IEEE

Antennas and Wireless Propagation Letters, vol. 8, pp. 1107 –1110, 2009.

[80] J.-F. Li, Q.-X. Chu, and T.-G. Huang, “A Compact Wideband MIMO Antenna with

Two Novel Bent Slits,” in IEEE Transactions on Antennas and Propagation, vol.

60, no. 2, pp. 482–489, Feb. 2012.

[81] C. Luo, J. Hong and L. Zhong, “Isolation Enhancement of a Very Compact UWB-

MIMO Slot Antenna with Two Defected Ground Structures,” in IEEE Antennas and

Wireless Propagation Letters, vol. 14, pp. 1766-1769, 2015.

[82] S. F. Jilani and A. Alomainy, “Millimetre-wave T-shaped MIMO antenna with

defected ground structures for 5G cellular networks,” in IET Microwaves, Antennas

& Propagation, vol. 12, no. 5, pp. 672-677, 18 4 2018.

95 [83] M. N. Hasan, S Chu, S. Bashir, “A DGS monopole antenna loaded with U-shape

stub for UWB MIMO applications,” Microwave and Optical Technology Letters,

vol. 61, no.9, pp. 1-9, 2019.

[84] H. Arun and M. Gulam Nabi Alsath. “Octagonal DGS based dual polarised ring-

shaped antenna for mimo communications,” International Journal of

Electronics, vol. 106, no. 5, pp.756-769, 2019.

[85] S. R. Pasumarthi, B. K. Jagadeesh and P. A. Mallikarjuna, “Design of Tri-Band

MIMO Antenna with Improved Isolation using DGS and Vias,” Wireless Personal

Communications, vol. 110, no. 3 pp. 1523-1532, 2020.

[86] S. Biswas et al., “High port isolation of a dual polarized microstrip antenna array

using DGS,” Journal of Electromagnetic Waves and Applications, pp. 1-14, 2020.

[87] J. Oh and K. Sarabandi, “Compact, Low Profile, Common Aperture Polarization,

and Pattern Diversity Antennas,” in IEEE Transactions on Antennas and

Propagation, vol. 62, no. 2, pp. 569-576, Feb. 2014.

[88] Y.S. Shin, S.O. Park, “A monopole antenna with a magneto-dielectric material and

its MIMO applications for 700MHz LTE band,” Microwave and Optical Technology

Letters, vol. 52, pp. 603–606, 2010.

[89] M. S. Sharawi, S. S. Iqbal and Y. S. Faouri, “An 800 MHz 2 X 1 compact MIMO

antenna system for LTE handsets,” in IEEE Transactions on Antennas and

Propagation, vol. 59, no. 8, pp. 3128-3131, Aug. 2011.

[90] L. Minz, R. Garg, “Reduction of mutual coupling between closely spaced PIFAs,”

Electronics Letters vol.46, no. 6, pp. 392–394, 2010.

96 [91] S. Chen, Y. Wang and S. Chung, “A Decoupling Technique for Increasing the Port

Isolation Between Two Strongly Coupled Antennas,” in IEEE Transactions on

Antennas and Propagation, vol. 56, no. 12, pp. 3650-3658, Dec. 2008.

[92] M. Krairiksh, P. Keowsawat, C. Phongcharoenpanich, “Two-probe excited circular

ring antenna for MIMO application,” Progress in Electromagnetics Research, vol.

97, pp. 417–431, 2009.

[93] J. OuYang, F. Yang, Z.M. Wang, “Reducing mutual coupling of closely spaced

microstrip MIMO antennas for WLAN application,” in IEEE Antennas and Wireless

Propagation Letters, vol. 10, pp. 310–313, 2011.

[94] A. R. Mallahzadeh, S. Es'haghi and A. Alipour, “Design of an E Shaped MIMO

antenna using IWO algorithm for wireless application at 5.8 GHz,” Progress in

Electromagnetics Research, vol. 97, pp. 417–431, 2009.

[95] X. Tang, K. Mouthaan, J.C. Coetzee, “Tunable decoupling and matching network

for diversity enhancement of closely spaced antennas,” in IEEE Antennas and

Wireless Propagation Letters, vol. 11, pp. 268–271, 2012.

[96] S. Ghosh, T. Tran and T. Le-Ngoc, “Miniaturized Four-Element Diversity PIFA,” in

IEEE Antennas and Wireless Propagation Letters, vol. 12, pp. 396-400, 2013.

[97] J. Lee, S. Kim and J. Jang, “Reduction of Mutual Coupling in Planar Multiple

Antenna by Using 1-D EBG and SRR Structures," in IEEE Transactions on

Antennas and Propagation, vol. 63, no. 9, pp. 4194-4198, Sept. 2015.

[98] A. Ramachandran, S. Valiyaveettil Pushpakaran, M. Pezholil and V. Kesavath, “A

Four-Port MIMO Antenna Using Concentric Square-Ring Patches Loaded With

CSRR for High Isolation,” in IEEE Antennas and Wireless Propagation Letters, vol.

15, pp. 1196-1199, 2016.

97 [99] H. Li, S. Sun, B. Wang and F. Wu, “Design of Compact Single-Layer Textile MIMO

Antenna for Wearable Applications,” in IEEE Transactions on Antennas and

Propagation, vol. 66, no. 6, pp. 3136-3141, June 2018.

[100]R. Li, P. Wang, Q. Zheng and R. Wu, “Compact microstrip decoupling and

matching network for two symmetric antennas,” in Electronics Letters, vol. 51, no.

18, pp. 1396-1398, 3 9 2015, doi: 10.1049/el.2015.1738.

[101]C. Xue, X. Y. Zhang, Y. F. Cao, Z. Hou and C. F. Ding, “MIMO Antenna Using

Hybrid Electric and Magnetic Coupling for Isolation Enhancement,” in IEEE

Transactions on Antennas and Propagation, vol. 65, no. 10, pp. 5162-5170, Oct.

2017, doi: 10.1109/TAP.2017.2738033.

[102] X. Zhou, R. Li, G. Jin, M.M. Tentzeris, “A compact broadband MIMO antenna for

mobile handset applications,” Microwave and Optical Technology Letters, vol. 53,

no. 12, pp. 2773–2776, 2011.

[103]S. Zhang, P. Zetterberg, S. He, “Printed MIMO antenna system of four closely-

spaced elements with large bandwidth and high isolation,” Electronics Letters, vol.

46, no. 15, pp. 1052–1053, 2010.

[104] G. Kang, Z. Du, K. Gong, “Compact broadband printed slot-monopole-hybrid

diversity antenna for mobile terminal,” IEEE Antennas and Wireless Propagation

Letters, vol. 10, pp. 159–162, 2011.

[105]C.H. See, R.A. Abd-Alhameed, Z.Z. Abidin, N.J. McEwan, P.S. Excell, “Wideband

printed MIMO/diversity monopole antenna for WiFi/WiMAX applications,” in

IEEE Transactions on Antennas and Propagation, vol. 60, no. 4, pp. 2028–2035,

2012.

98 [106]Z. Tang, et al., “Design of a compact UWB-MIMO antenna with high isolation and

dual band-notched characteristics,” Journal of Electromagnetic Waves and

Applications, vol. 34, no. 4, pp. 500-513, 2020.

[107]Weiye Li, Wenbin Lin, Guangli Yang, “A Compact MIMO Antenna System Design

with Low Correlation from 1710MHz to 2690MHz,” Progress In Electromagnetics

Research, vol. 144, pp. 59-65, 2014.

[108]S. Shoaib, I. Shoaib, N. Shoaib, X. Chen and C. G. Parini, “MIMO Antennas for

Mobile Handsets,” in IEEE Antennas and Wireless Propagation Letters, vol. 14,

pp. 799-802, 2015.

[109]X. Zhao, S. P. Yeo and L. C. Ong, “Planar UWB MIMO Antenna with Pattern

Diversity and Isolation Improvement for Mobile Platform Based on the Theory of

Characteristic Modes,” in IEEE Transactions on Antennas and Propagation, vol.

66, no. 1, pp. 420-425, Jan. 2018.

[110]S. Rajkumar, A. Anto Amala and K. T. Selvan, “Isolation improvement of UWB

MIMO antenna utilising molecule fractal structure,” in Electronics Letters, vol.

55, no. 10, pp. 576-579, 16 5 2019.

[111]Y. Ding, Z. Du, K. Gong, Z. Feng, “A novel dual-band printed diversity antenna for

mobile terminals,” IEEE Transactions on Antennas and Propagation, vol. 55, pp.

2088–2096, 2007.

[112]S. Zhang, B. K. Lau, Y. Tan, Z. Ying and S. He, “Mutual Coupling Reduction of

Two PIFAs with a T-Shape Slot Impedance Transformer for MIMO Mobile

Terminals,” in IEEE Transactions on Antennas and Propagation, vol. 60, no. 3, pp.

1521-1531, March 2012.

99 [113]R. Addaci, A. Diallo, C. Luxey, P. Le Thuc and R. Staraj, “Dual-Band WLAN

Diversity Antenna System with High Port-to-Port Isolation,” in IEEE Antennas and

Wireless Propagation Letters, vol. 11, pp. 244-247, 2012.

[114]X. Zhou, X. Quan and R. Li, “A Dual-Broadband MIMO Antenna System for

GSM/UMTS/LTE and WLAN Handsets,” in IEEE Antennas and Wireless

Propagation Letters, vol. 11, pp. 551-554, 2012.

[115]Q. Rao and K. Wilson, “Design, Modeling, and Evaluation of a Multiband

MIMO/Diversity Antenna System for Small Wireless Mobile Terminals,” in IEEE

Transactions on Components, Packaging and Manufacturing Technology, vol. 1,

no. 3, pp. 410-419, March 2011.

[116]H. S. Singh, B. R. Meruva, G. K. Pandey, P. K. Bharti, and M. K. Meshram, “Low

mutual coupling between MIMO antennas by using two folded shorting strips,”

Progress in Electromagnetics Research B, vol. 53, pp. 205–221, Jul. 2013.

[117]R. Addaci et al., “Dual-Band WLAN Multiantenna System and Diversity/MIMO

Performance Evaluation,” in IEEE Transactions on Antennas and Propagation, vol.

62, no. 3, pp. 1409-1415, March 2014.

[118]H. S. Singh et al., “A compact dual-band diversity antenna for WLAN applications

with high isolation,” Microwave and optical technology letters, vol. 57, No. 4, 906-

912, 2015.

[119]X. Luo, J. Yuan and K. Chen, “Compact and Low Profile MIMO Antenna for Dual-

WLAN-Band Access Points,” Progress In Electromagnetics Research Letters, vol.

67, 97-102, 2017.

100 [120] K. V. Babu and B. Anuradha, “Design of Multi-band Minkowski MIMO Antenna

to reduce the mutual coupling,” Journal of King Saud University-Engineering

Sciences, 2018.

[121]Y. Ou, X. Cai and K. Qian, “Two-element compact antennas decoupled with a

simple neutralization line,” Progress in Electromagnetics Research., vol. 65, pp.

63-68, 2017.

[122]X. Wu, G. Han and C. Feng, “Dual Frequency MIMO Antenna with Neutralization

Line,” 2019 Cross Strait Quad-Regional Radio Science and Wireless Technology

Conference (CSQRWC), Taiyuan, China, 2019, pp. 1-2, doi:

10.1109/CSQRWC.2019.8799229.

[123]C. Luo, et al.: “A decoupling method between two tri-band antennas for

WLAN/WiMAX applications,” IEICE Electron. Express, vol. 14, 2017, 20170354

(DOI: 10.1587/elex.14.20170354).

[124]R. Liu, X. An, H. Zheng, M. Wang, Z. Gao and E. Li, “Neutralization Line

Decoupling Tri-Band Multiple-Input Multiple-Output Antenna Design,” in IEEE

Access, vol. 8, pp. 27018-27026, 2020, doi: 10.1109/ACCESS.2020.2971038.

[125]A. Dkiouak, A. Zakriti, M. E. Ouahabi and A. Mchbal, “Design of a two symetrical

F-shaped MIMO antenna for Wi-MAX and WLAN applications,” 2019

International Conference on Wireless Technologies, Embedded and Intelligent

Systems (WITS), Fez, Morocco, 2019, pp. 1-6, doi: 10.1109/WITS.2019.8723859.

[126]P. Liu, D. Sun, P. Wang and P. Gao, “Design of a Dual-Band MIMO Antenna with

High Isolation for WLAN Applications,” Progress In Electromagnetics Research,

vol. 74, pp.23-30, 2018.

101 [127]R.Addaci, A. Diallo, C. Luxey, and P.L. Thuc, and R. Staraj, “Dual-band WLAN

diversity antenna-system with high port-to-port isolation,” in IEEE Antennas,

Wireless Propagation Letters, vol. 11, 2012, pp. 244–247.

[128]M. S. Sharawi, A. B. Numan and D. N. Aloi, “Isolation Improvement in a Dual-

Element MIMO Antenna System Using Capacitively Loaded Loops,” Progress In

Electromagnetics Research, vol. 134, pp. 247-266, 2013.

[129]W. Wu et al, “A compact multiband MIMO antenna for IEEE 802.11 a/b/g/n

applications,” Progress In Electromagnetics Research Letters, vol. 84, pp. 59–65,

2019.

[130]Y. Yao, X. Wang, X. Chen, J. Yu and S. Liu, “Novel Diversity/MIMO PIFA

Antenna With Broadband Circular Polarization for Multimode Satellite

Navigation,” in IEEE Antennas and Wireless Propagation Letters, vol. 11, pp. 65-

68, 2012, doi: 10.1109/LAWP.2012.2183335

[131]B. P. Chacko, G. Augustin and T. A. Denidni, “Compact uni-planar antenna with

polarization diversity for UWB application in portable devices,” 2014 IEEE

Antennas and Propagation Society International Symposium (APSURSI),

Memphis, TN, 2014, pp. 1809-1810, doi: 10.1109/APS.2014.6905231.

[132]S. Tripathi, A. Mohan and S. Yadav, “A Compact Koch Fractal UWB MIMO

Antenna With WLAN Band-Rejection,” in IEEE Antennas and Wireless

Propagation Letters, vol. 14, pp. 1565-1568, 2015, doi:

10.1109/LAWP.2015.2412659.

[133]H.S. Singh and M.K. Meshram, “Printed Monopole Diversity Antenna for USB

Dongle Applications,” Wireless Personal Communications, vol. 86, no. 2, pp.771-

787, 2016.

102 [134]S. W. Cheung, Q. Li, D. Wu, C. Zhou and B. Wang, “Defected ground structure

with two resonances for decoupling of dual-band MIMO antenna,” 2017 IEEE

International Symposium on Antennas and Propagation & USNC/URSI National

Radio Science Meeting, San Diego, CA, 2017, pp. 1645-1646, doi:

10.1109/APUSNCURSINRSM.2017.8072865.

[135]S. R. Thummaluru and R. K. Chaudhary, “Mu-negative metamaterial filter-based

isolation technique for MIMO antennas,” in Electronics Letters, vol. 53, no. 10, pp.

644-646, 11 5 2017, doi: 10.1049/el.2017.0809.

[136]P. Mondal, D. Dhara and A. R. Harish, “A Wideband Slotted D-Shaped Monopole

MIMO Antenna With Triple Branch Neutralization Line,” 2018 IEEE Indian

Conference on Antennas and Propagation (InCAP), Hyderabad, India, 2018, pp. 1-

4, doi: 10.1109/INCAP.2018.8770733.

[137]Y. Zhang, Y. Jiao, M. Yang and B. Zhao, “A tri-band notched Ultra-wideband

MIMO Antenna,” 2019 International Symposium on Antennas and Propagation

(ISAP), Xi'an, China, 2019, pp. 1-3.

[138]S. Luo, Y. Li, Y. Xia, and L. Zhang, “A low mutual coupling antenna array with

gain enhancement using metamaterial loading and neutralization line structure,”

Applied Computational Electromagnetics Society, vol. 34, no. 3, pp. 411–418, Mar.

2019.

[139] M. Manteghi and Y.Rahmat-Samii, “Multiport characteristics of a wideband

cavity-backed annular patch antenna for multi-polarization operations,” IEEE

Transactions on Antennas and Propagation, vol.53, no.1, pp. 466- 474, 2005.

103 [140]S. Blanch, J. Romeu, I. Corbella, “Exact representation of antenna system diversity

performance from input parameter description,” IET Electronic Letters, vol. 39, No.

9, pp. 707-707, May 2003.

[141]S. C. K. Ko and R. D. Murch, “Compact integrated diversity antenna for wireless

communications,” in IEEE Transactions on antenna and propagation, vol. 49, no.

6, pp. 954–960, Jun. 2001.

[142]R. G. Vaughan and J. B. Andersen, “Antenna diversity in mobile communications,”

in IEEE Transactions on Vehicular Technology, vol. VT-36, pp. 147–172, Nov.

1987.

[143]H. Shin and J. H. Lee, “Capacity of multiple-antenna fading channels: spatial fading

correlation, double scattering, and keyhole,” in IEEE Transactions on Information

Theory, vol. 49, no. 10, pp. 2636-2647, Oct. 2003, doi: 10.1109/TIT.2003.817439.

[144]R. Vaughan and J. Bach Andersen, “Channels, Propagation and Antennas for

Mobile Communications,” IET Digital Library, 2003.

104

105 CHAPTER THREE

Mutual Coupling Reduction of Dual-Band Printed MIMO Antenna Using Neutralization Line Technique and Defected Ground Structures

The main potential objective of MIMO antenna systems is to enhance reliability and improve channel capacity in wireless communication systems. Consequently, they have generated great and radical innovations in applications for wireless communications systems research areas [1]. A significant design consideration with MIMO antennas is the ability to minimize spatial correlation factor, and the major contribution effects resulting from the coupling of the radio frequency (RF) electromagnetic (EM) interactions that exist between closely adjacent radiating elements: these factors can largely define the system performance of the MIMO antenna. High isolation can result in higher antenna efficiencies and lower correlation coefficients [2]. Many approaches have been presented in the literature that aim to improve the isolation over a multi-band operation regime for MIMO antennas. However, no significant contribution has been made to improve the isolation in the relevant bands using a hybrid approach.

This chapter presents a new compact multiband MIMO antenna system, which can operate at WLAN bands (2.4/5.2/5.8 GHz) and achieve good isolation in the relevant bands. It is composed of two uni-planar double T-shaped monopole antennas placed symmetrically on the same substrate. The coupling at the upper frequency band is minimized using the defected ground structure method while the coupling at the lower frequency band is reduced using the neutralization line technique. The antenna

106 performance characteristics are reported in terms of scattering parameters, ECC, TARC,

CCL, DG, antenna peak gain, antenna efficiency and antenna radiation patterns.

3.1 Single Antenna Design

In this design, a dual-band antenna with two double T-shaped radiating elements is modelled and simulated, aimed at fulfilling the requirements of WLAN applications. As a substrate material for the proposed design, FR4 glass-reinforced epoxy laminate with

1.6 mm thickness, 4.3 relative permittivity, and 0.025 loss tangent was used. The overall dimensions were 35.516.51.6 mm3. The antenna ground plane was printed on the bottom layer of the substrate. The geometry of the proposed antenna is illustrated in Fig.

3.1 while all the optimized dimensions are listed in Table 3.1.

Initially, a compact double T-shaped antenna was modelled to work at dual frequency bands. The simple approach to design an antenna with a dual-band operation is by obtaining two elements with different lengths inside the antenna structure, so that each element will resonate at the desired frequency by matching its length with a quarter of the wavelength of the operating frequency in the mixed air-substrate medium [3].

Therefore, the upper-frequency band is controlled by the shorter T-shaped arm while the lower frequency band is controlled by the larger T-shaped arm. The variation of the reflection coefficient versus frequency of the proposed antenna is plotted in Fig. 3.2. It can be clearly seen that the antenna covers the two required bands with the first band ranging from 2.37 GHz to 2.8 GHz and the second band ranging from 5.15 GHz to 6.25

GHz.

107

To evaluate the effect of the two arms of the proposed single antenna, simulations were performed to study the effect of changing the length of the upper arm (arm1) only and keeping other parameters at the same values. The obtained reflection coefficient results are shown in Fig. 3.3, where it can be seen that the first frequency band is shifted to 2.5

GHz by increasing the length of the upper arm and at the same time the second frequency band is expanded while the impedance matching is reduced. Moreover, the other factor which is the length of the lower arm (arm2) is also investigated to determine its effect on the reflection coefficient of the proposed antenna. Fig. 3.4 shows the obtained S11 results with different values of the length of the lower arm. It can be clearly seen that there is no effect on the first band by increasing the length of the small arm, while the resonant frequency of the second band is shifted to toward 5.5 GHz.

Fig. 3.1. Geometrical configurations of the proposed single antenna.

Table 3.1. Single antenna parameter values.

108 Parameters Value Parameters Value Parameters Value Parameters Value

h1 7.8mm h5 2.5mm L1 16mm L6 5.75mm

h2 4mm h6 17mm L3 4.75mm L7 1.5mm

h3 2.5mm h7 15mm L4 6.75mm L8 2.6mm

h4 17mm h8 1mm L5 5.25mm L9 7.25mm

Fig. 3.2. Simulated reflection coefficients S11 of the proposed single antenna.

Fig. 3.3. Variation of simulated S11 versus frequency with different length of arm1.

109

Fig. 3.4. Variation of the simulated S11 versus frequency with different length of arm2.

The effect of tapering the ground plane by cutting two triangles from its upper corners as shown in Fig. 3.5 was investigated. The reflection coefficient characteristics for various values of the height (r) of the cut triangle are shown in Fig. 3.6. It can be seen that the taper has little effect on the first band as it shifts the centre frequency slightly to lower values while it increases the bandwidth of the second band. As can be seen, the case for r=2.5 mm gives the best results among the investigated values. The reason for this effect can be attributed to the fact that the tapering increases the separation between the ground plane and lower T-shaped arm thus allowing better current distribution along this edge.

Fig. 3.5. Back view of the proposed single antenna showing the two defected areas.

110

Fig. 3.6. Simulated reflection coefficients S11 of the proposed single antenna with different values of r (mm).

To obtain better insight into the dual band operation of the antenna, the surface current distribution across the arms was found using the CST Microwave Studio software. The obtained current distributions at two selected frequencies are shown in Fig. 3.7. It can be seen that at the lower frequency band the upper arm has a rich current distribution, while at the upper frequency band the lower arm has richer surface current. This indicates that the upper arm controls the lower band frequency, while the lower arm influences the upper band frequency.

It is beneficial to calculate the antenna gain and efficiency for the proposed antenna.

These parameters can give extra information about the antenna performance. The peak gain and radiation efficiency are evaluated using the CST package. Fig. 3.8 shows that the peak gains of the proposed antenna are 1.85 dBi and 2.5 dBi at the lower and upper frequency bands, respectively. On the other hand, the radiation efficiencies of the

111 proposed antenna are 97% and 70% at the centre frequencies of the lower and upper bands respectively, as shown in Fig 3.8. Otherwise, there is a dip in antenna gain and efficiency at 4 GHz: this is due to the high mismatch between antenna ports at this frequency, as shown in Fig. 3.2.

Fig. 3.7. Surface current distributions of the proposed single antenna at two different frequencies; (a) 2.4 GHz, (b) 5.5 GHz.

Fig. 3.8. Variation of calculated peak gain and efficiency versus frequency for the proposed single antenna.

112 3.2 2 2 MIMO Antenna Design

In this section, a 22 MIMO antenna was adapted from the single antenna design for

WLAN applications. The MIMO antenna was modelled using the same FR4 substrate.

The overall size of the antenna was 3633.51.6 mm3. The inter-element distance between the radiators was kept at 3 mm, which is equivalent to 0.024 λo. On the other hand, the centre to centre separation was kept at 16.5 mm. The antennas share the same ground plane in the bottom layer of the substrate. Moreover, a new approach that combines DGS and neutralization line decoupling techniques was applied to enhance the antenna isolation. NL with a length of 16.5 mm was constructed to connect the two radiating elements. The length of the NL is equivalent to g/4 at the isolation frequency

2.4 GHz. Furthermore, the DGS decoupling approach was applied by creating defected areas in the corners of the ground plane to reduce the coupling as well as enlarge the bandwidth at the upper frequency band. The schematic structure of the proposed antenna is shown in Fig. 3.9 and all the optimized dimensions are summarized in Table 3.2.

To obtain better insight into the contributions of the long and short T-shaped arms within the proposed MIMO antenna, the surface current distribution across the radiating elements was simulated using the CST software at the resonant frequency of the two required bands by exciting port one and terminating port two, as shown in Fig. 3.10. It can be observed that the distribution of the surface current at 2.4 GHz is mainly located along the edges of the upper arm as well as the feed line. This indicates that the lower frequency is controlled by the long arm. On the other hand, the distribution of the surface current at 5.5 GHz is mainly located across the edges of the lower arm as well as the feed line. This demonstrates that the upper frequency is controlled by the short arm [4].

113

Fig. 3.9. The configuration of the proposed antenna; (a) Antenna with NL (b) Antenna without NL.

Table 3.2. MIMO Antenna parameter values.

Parameters Value Parameters Value Parameters Value Parameters Value

h1 7.8mm h5 2.5mm L1 16mm L5 5.25mm

h2 4mm h6 17mm L2 16.5mm L6 5.75mm

h3 2.5mm h7 15mm L3 4.75mm L7 1.5mm

h4 17mm h8 1mm L4 6.75mm L8 2.6mm

L9 7.25mm

114

Fig. 3.10. Surface current distributions of the proposed MIMO antenna without NL at the resonant frequency of the two required bands; (a) 2.4 GHz. (b) 5.5 GHz.

3.3 Neutralization Line

The neutralization line technique can be classified as one of the effective methods that are used to improve the isolation in MIMO antenna systems, as illustrated in [5-11].

According to Chebihi et al. in [9], the main function of the NL is to pass the electromagnetic (EM) wave from one radiating element to the other one at a certain point inside the MIMO antenna structure via a metallic slit or lumped element. This EM wave should have an opposite coupling to reduce the main coupling in the antenna system at a specific frequency band [12]. The critical issue associated with this technique is the selection of the connection point. The position of the point inside the radiating structure should have low impedance and high current [13]. Two parameters should be taken into account in the design of the NL; the length and the width. These parameters have a great effect on the results for mutual coupling. Recently, the neutralization line technique was adopted by several mobile phone companies, for example, Samsung [14] and LG

Electronics [15], as a very effective mechanism for solving isolation problems in MIMO antenna systems. 115

Several types of neutralization line structures have been used in the literature, such as thin printed neutralization lines [16], a pair of crossed neutralization lines [17], LC matching network and NL [18] and neutralization lines between ground planes [19].

3.3.1 S-Parameters of the Proposed MIMO Antenna with and without NL

The S-parameters of the antenna with and without the neutralization line have been simulated to verify the effectiveness of the NL. Fig. 3.11 shows the simulated result of the reflection coefficients S11. It can be clearly observed that there is a shift to the right side (toward 2.5 GHz) in the first band of operation and a reduction in the bandwidth of the same band. On the other hand, there is no effect on the second band after inserting the

NL. Therefore, the same band has been achieved.

The simulated coupling with and without NL is illustrated in Fig. 3.12. It can be clearly seen that the mutual coupling around 2.5 GHz without NL is higher than -10 dB. By inserting the neutralization line, the mutual coupling between the elements has been decreased in that band: the simulated isolation is then nearly 27 dB at 2.5 GHz and 13 dB at 5.2 and 5.8 GHz, as shown in Fig. 3.12.

116

Fig. 3.11. Simulated reflection coefficients S11 with and without neutralization line for the proposed MIMO antenna.

Fig. 3.12. Simulated transmission coefficients S21 with and without the neutralization line for the proposed MIMO antenna.

117 Moreover, the surface current distribution of the proposed antenna was used to analyse the contributions of the neutralization line properly. Two cases were studied at 2.4 GHz: the first case was the proposed antenna excluding the neutralization line while the second case was the proposed antenna including the neutralization line. This study is performed by exciting port 1 and terminating port 2 with a matched load. Fig. 3.13 demonstrates the surface current distribution with and without NL. It can be clearly seen that the induced current in the feeding line port 2 is high in the case of the absence of the neutralization line. On the other hand, the existence of the neutralization line will introduce a new current path which generates an additional coupling to reduce the original coupling, as shown in Fig. 3.13-a [4].

Fig. 3.13. Surface current distributions of the proposed MIMO antenna at 2.4 GHz; (a)

With NL, (b) Without NL.

3.3.2 Parametric Study

In order to clarify the effects and obtain the optimized values of the proposed design, parametric studies of the position and width of the neutralization line based on Fig. 3.14 were carried out.

118

Fig. 3.14. Configuration of the proposed MIMO antenna showing the height (hn) and the width (wn) of the NL.

3.3.2.1 The Effect of the Position of the Neutralization Line on the S-Parameters

To validate the effect of the position of the neutralization line, the simulated S-parameters of the antenna with different positions (hn) of the NL were investigated, as shown in Fig.

3.15 and Fig. 3.16 respectively. Fig. 3.15 shows that by raising the position of the NL, the first band is shifted toward the 2.5 GHz band while the beginning of the second band is shifted backward. On the other hand, raising the position of the NL leads to a decrease in the mutual coupling between the elements in the first band and there are no effects on the second band. Tables 3.3 and 3.4 summarize the values for the two figures in order to clarify the results.

119

Fig. 3.15. The simulated reflection coefficient of the proposed MIMO antenna with different positions for the NL.

Fig. 3.16. The simulated transmission coefficient S21 of the proposed MIMO antenna with different positions of the NL.

120

Table 3.3. The effect of the position of the NL on the reflection coefficients.

hn=22 mm hm=23 mm hm=24 mm hm=25 mm

f= 2.5 GHz -11.1 dB -12.8 dB -16.7 dB -16.52 dB

f=5.2 GHz -16.58 dB -15.24 dB -15.8 dB -15.7 dB

f=5.8 GHz -13 dB -12.48 dB -12.29 dB -12.32 dB

Table 3.4. The effect of the position of the NL on the transmission coefficients.

hn=22 mm hm=23 mm hm=24 mm hm=25 mm

f= 2.5 GHz -8.3 dB -9.5 dB -12.2 dB -18.7 dB

f=5.2 GHz -13.19 dB -13.9 dB -13.99 dB -13.64 dB

f=5.8 GHz -13.76 dB -13.45 dB -13.28 dB -12.68 dB

3.3.2.2 The Effect of the Width of the NL on the S-parameters of the Proposed Antenna

The width of the NL was changed according to Fig. 3.14. Fig. 3.17 and Fig. 3.18 represent the reflection coefficients (S11) and mutual coupling values (S21) respectively. It can be clearly seen that, by increasing the width of the NL, the reflection coefficients are

121 decreased and there are no effects on the two frequency bands. On the other hand, the mutual coupling values decrease by increasing the width of the NL for the first band and there is a small effect on the second band. The summarized values of the two figures are listed in Table 3.5 and Table 3.6.

Fig. 3.17. Simulated reflection coefficient S11 of the proposed MIMO antenna with different widths of the NL.

Fig. 3.18. Simulated transmission coefficient S21 of the proposed MIMO antenna with different widths of the NL.

122

Table 3.5. The effect of the width of the NL on the reflection coefficients.

wn=0.25 mm wn=0.5 mm wn=0.75 mm wn=1 mm

f=2.5 GHz -18.7 dB -16.33 dB -13.9 dB -12.6 dB

f=5.2 GHz -15.7 dB -15.69 dB -15.6 dB -15.6 dB

f=5.8 GHz -12.32 dB -12.24 dB -12.1 dB -12.1 dB

Table 3.6. The effect of the width of the NL on the transmission coefficients.

wn=0.25 mm wn=0.5 mm wn=0.75 mm wn=1 mm

f=2.5 GHz -24 dB -22.5 dB -18.5dB -16 dB

f=5.2 GHz -13.69 dB -13.66 dB -13.68 dB -13.68 dB

f=5.8 GHz -12.8 dB -12.7 dB -12.7 dB -12.7 dB

3.4 Defected Ground Structure

The concept of Defected Ground Structures is derived from Photonic Bandgap Structures

(PBG) in electromagnetics. In general, for RF and microwave applications, PBGs are expressed as Electromagnetic Bandgap (EBG). These periodic structures are used to prevent EM waves from passing through them over a frequency band called the stopband and to allow EM waves to pass through them over a frequency band called the passband

123 [20]. Defected ground structures refer to specific configurations, called unit cells. These unit cells are etched out on the ground plane of a MIMO antenna or PCB in the form of a single defect or periodic geometries to prevent EM waves from passing through the substrate over a specific frequency band. The DGS slots have various configurations and sizes with different frequency responses. The existence of DGS in a printed MIMO antenna system will help to stop the movement of the surface current in the ground plane over the defected region [21]. This will lead to a reduction in the mutual coupling between

MIMO antennas. Additionally, DGS can improve several other parameters in MIMO antennas such as bandwidth, antenna gain and cross-polarization [22].

3.4.1 S-Parameters of the Proposed Antenna with DGS

The investigations started with assessment of the effect of defecting the ground plane by cutting three different areas in the middle, right side and left side of the upper part of the ground plane, as shown in Fig. 3.19. The S-parameter characteristics for various values of the height of the cut area (r) are shown in Fig. 3.20 and Fig. 3.21 respectively. It can be seen that increasing the height of the cut area (r) has little effect on the reflection coefficient at the lower frequency band, as exhibited by the slight shift in the centre frequency, while it increases the bandwidth of the reflection coefficient significantly at the second band. Furthermore, the case for r = 2 mm produces the best results among the investigated ones, as shown in Fig. 3.20. On the other hand, increasing the height of the cut area (r) does not affect the isolation of the first band and it shows good correspondence with various values of (r), while the isolation in the second band is improved by increasing the parameter (r) and the best result among the investigated ones is in the case of r = 2

124 mm, where the isolation is more than 13 dB in the whole second band, as shown in Fig.

3.21.

Fig. 3.19. Back view of the proposed MIMO antenna showing the defected areas.

Fig. 3.20. Simulated reflection coefficients of the proposed MIMO antenna with different values of r (mm).

125

Fig. 3.21. Simulated transmission coefficients of the proposed MIMO antenna with different values of r (mm).

3.5 Evaluation of Antenna Efficiency and Gain

It is crucial to evaluate the antenna efficiency and gain of the proposed MIMO antenna system after using the neutralization line. Therefore, the antenna radiation efficiency and peak gain of the proposed MIMO antenna with decoupling were simulated and compared with the values of the single antenna and the MIMO antennas without decoupling.

The proposed MIMO antenna with neutralization line achieves an efficiency equal to 97% over the first band, an increase of 7% compared to the efficiency of the single antenna.

On the other hand, the efficiency of the proposed MIMO antenna with neutralization line is 80%, an improvement of 3% compared to the efficiency of the single antenna. The obtained efficiencies versus frequency are shown in Fig. 3.22. Moreover, Fig. 3.23 explains the simulated peak gain of the proposed MIMO antenna over the interval from

2 to 6.5 GHz. At the lower frequency band, the peak gain of the proposed antenna with

126 NL is 2 dBi, an improvement of 1.2 dBi compared to the MIMO antenna without NL, while the peak gain increased in the second band and it achieves 3 dBi at 5.5 GHz, a 0.8 dBi enhancement compared to the peak gain of the MIMO antenna without NL.

Fig. 3.22. Variation of calculated efficiency versus frequency for the single antenna, the

MIMO antenna without decoupling and MIMO antenna with decoupling.

Fig. 3.23. Variation of calculated peak gain versus frequency for the single antenna, the

MIMO antenna without decoupling and MIMO antenna with decoupling.

127 3.6 Results and Discussions

Simulated and measured outcomes of the proposed MIMO antenna array with neutralization line are presented here to verify the antenna performance. Distinguishing features such as high isolation between the radiating elements, low values for envelope correlation coefficient and channel capacity loss, TARC value close to -10 dB within the desired frequency band and omnidirectional radiation pattern, indicate the potential of the proposed MIMO antenna. The designed antenna was prototyped using the same substrate and dimensions used in the simulation as shown in Fig. 3.24. The comparisons between the simulated and the measured results in terms of S-parameters are plotted in Fig. 3.25 and Fig 3.26 respectively. It can be clearly seen that there is a reasonable agreement between results with some tolerance that can be attributed to reflections from the SMA connector and the objects near the antenna during the tests. The impedance bandwidth of the antenna covers the operating frequency bands from 2.4 to 2.7 GHz and from 4.4 to

6.7 GHz for reflection coefficient |S11| < -10 dB. These achieved bandwidths fulfil the requirements of the WLAN applications. On the other hand, the results of S21 for the whole band (2-6.7 GHz) are better than -10 dB and it reaches to -30 dB around 2.4 and

5.5 GHz.

Fig. 3.24.The fabricated MIMO antenna design with NL (a) Top view, (b) Back view.

128

Fig. 3.25. Simulated and measured S11 of the proposed MIMO antenna with decoupling methods.

Fig. 3.26. Simulated and measured S21 of the proposed MIMO antenna with decoupling methods.

129

Generally, the envelope correlation coefficient is a significant factor in MIMO antenna systems, as explained in Chapter Two. The evaluation of the ECC of a MIMO antenna could be obtained through two different approaches. The first method depends on the far- field radiation pattern of the antenna [23, 24] and the second method uses the S- parameters of the antenna [25]. The simulated and measured ECC of the proposed MIMO antenna were calculated using equation (3.1) which is based on the second method and the results are illustrated in Fig. 3.27. It can be clearly seen from the figure that the value of the ECC is less than 0.002 in two bands, which is significantly less than 0.3, the maximum allowed for MIMO antenna applications [5].

|푆 ∗ 11푆12 + 푆 ∗ 21푆22|

퐸퐶퐶 = 2 2 (3.1) (1 ― |푆 11| ― |푆21|²) (1 ― |푆22| ― |푆 12|²)

Fig. 3.27. Simulated and measured ECC of the proposed MIMO antenna with NL.

130 The next metric that is calculated to assess the proposed MIMO antenna performance is the channel capacity loss. In any conventional wireless system, increasing the number of utilised antennas will lead to a linear increase in the channel capacity, without raising the bandwidth or transmitted power [26, 27]. In this design, the simulated and measured channel capacity loss are compared and plotted in Fig 3.28, depending on the equation

(2.22). It can be easily seen that the CCL is less than 0.5 bit/s/Hz for the two frequency bands and there is fair agreement between the calculated and measured results. These consequences mean that a good impedance match with low mutual coupling between the two antenna elements leads to a low capacity loss in the MIMO channel.

Fig. 3.28. Simulated and measured CCL of the proposed MIMO antenna decoupling methods.

The total active reflection coefficient is another significant parameter in the MIMO antenna system which is used to properly characterize the bandwidth and efficiency of

131 the system. TARC is defined as the result of dividing the square root of the total reflected power to the square root of the total incident power [23]. Equation (3.2) is used to evaluate the TARC value in a two-port MIMO antenna system [28].

푗휃 푗휃 |푆11 + 푆12푒 |² + |푆21 + 푆22푒 |² 푡 (( ) ( )) 휏푎 = (3.2) 2

The TARCs of the present design are calculated based on equation (3.2) and are plotted in Fig. 3.29 and Fig. 3.30 respectively. Fig. 3.29 shows TARC curves with different phases. These curves cover the phase range from 0o to 180o with phase steps of 30o. The average value of TARC is shown in Fig 3.30. It can be clearly seen that the two bands appeared with TARC value less than -10 dB.

Fig. 3.29. Calculated TARC with a different phase of the proposed MIMO antenna with decoupling methods.

132

Fig. 3.30. The calculated average value of TARC for the proposed MIMO antenna with decoupling methods.

Diversity Gain is another important factor that affects the performance of a MIMO antenna system. This metric can be defined as the amount of improvement delivered by the diversity antenna system compared to a single antenna system in one diversity channel

[23]. In this section, the simulated and measured DG of the proposed MIMO antenna is presented in Fig. 3.31 and it can be clearly seen that the DG values for the two frequency bands are very close to the ideal value of DG =10 dB.

Finally, the prototyped design of the proposed antenna was tested in terms of far-field radiation patterns at four different frequencies in the anechoic chamber of the University of Bradford. These patterns were measured in the two planes of XZ and YZ, in the case of exciting port 1 and terminating port 2. An elevation-over-azimuth positioner was used

133 by synchronizing the elevation axis with the MIMO antenna’s coordinate system at (θ =

0o). Therefore, cuts at constant Ø will be generated by the azimuth drive. A broadband horn (EMCO type 3115) was used as a transmitting antenna and a 4m distance was kept between the transmitter and the antenna under test. The measurements were carried out by rotating the azimuth positioner from θ= -180o to 180o at increments of 5o. These patterns were measured in the two planes of XZ and YZ, in the case of exciting port 1 and terminating port 2. The practical results are shown in Fig. 3.32 and they show that at the three frequencies, the far-field radiations achieve stable omnidirectional patterns.

Fig. 3.31. Measured and simulated diversity gain of the proposed MIMO antenna decoupling methods.

134

Fig. 3.32. Simulated and measured radiation patterns of the proposed antenna for two planes [(1 and 2) y-x plane and (3 and 4) z-x plane] at (a) 2.4 GHz (b) 5.2 GHz and (c)

5.8 GHz. Port 1 is excited and port 2 is terminated. Solid curve: simulated results, dashed curve: measured results, “1 and 3” co-polar components and “2 and 4” cross-polar components.

For further evaluation of the volumetric radiation patterns, the three-dimensional representation of the radiated field for the proposed antenna was calculated at different frequency bands and is shown in Fig. 3.33. This figure gives more appreciation of the field shape as compared to those of the 2-D representations.

135

Fig. 3.33. The simulated 3-D pattern of the proposed MIMO antenna at (a) 2.4 GHz (b)

5.2 GHz and (c) 5.8 GHz.

3.7 Comparison with Published Works

It is instructive to compare the proposed MIMO antenna with several examples of published data from other researchers. All the antennas in this comparison are monopole antennas to ensure a like for like comparison and they use either neutralization line or

DGS as coupling reduction techniques. This comparison is not comprehensive but it includes a fair set of representatives for the state of the art of this technology. The comparison depends on the centre frequency, the isolation between the elements, the overall electrical size of the antenna, the separation distance between the radiating

136 elements in terms of λo (where λo is the free space wavelength at the lower centre frequency), antenna gain, antenna efficiency and ECC. The summary of this comparison is listed in Table 3.7. It can be noticed that the principal dimensions of the proposed antenna are smaller than for the other antennas. The isolation of the radiating elements is shown to provide a good value compared to the distance between them. High efficiency with a low correlation coefficient (compared with other designs) have also been obtained.

It is important to mention here that most of the other published research works do not calculate all the influencing parameters. This gives a lack of information to determine the performance of the MIMO antenna.

Table 3.7. Comparison with other works.

Centre Peak Antenna Isolation Separation Ref. Frequency Electrical Size Gain Efficiency ECC (dB) distance (GHz) (dBi) (%) 0.99λo×0.44λo [30] 3.3 >17 0.206 λo 2.5 <0.01 ×0.0087λo 0.47λo×0.44λo [31] 4.05 >22 0.0297 λo 2.8 90 <0.1 ×0.01λo 0.73λo×0.9λo [32] 8.275 >20 0.014 λo 2 ≤0.1 ×0.044λo 1.95λo×0.97λo [33] 7.34 >15 0.376 λo <0.015 ×0.038λo 2.9 0.61λo×0.29λo [34] 4.3 >15 0.087 λo 0.0001 7.7 ×0.015 λo

This 2.55 0.306λo×0.28λo 2 97 >15 0.0255 λo <0.002 work 5.5 ×0.013λo 3 80

3.8 Conclusions

A compact dual-band MIMO antenna operating at the WLAN bands (2.4/5.2/5.8 GHz) has been designed. The schematic structure of this antenna consists of two double T- shaped radiating elements with a neutralization line connecting them. In this design, the coupling at the upper frequency band is minimized using the defected ground structure

137 method while the coupling at the lower frequency band is reduced using the neutralization line method. The optimum value of isolation has been obtained by optimizing the dimensions of the decoupling methods. The antenna has achieved bandwidths from 2.4 to 2.7 GHz and from 4.4 to 6.7 GHz for reflection coefficient |S11| < -10 dB. On the other hand, the results of S21 for the whole band (1-6 GHz) are better than -15 dB and it has reached to -30 dB around 2.4 and 5.5 GHz. The envelope correlation coefficient has also been evaluated and found to be less than 0.002 in two bands. Channel capacity loss is another parameter which has been calculated and presented. The obtained value of CCL is less than 0.5 bps/s/Hz in the two required bands. The simulated and measured average

TARCs have been also illustrated. It has been observed that the two bands appeared with

TARC values less than -10 dB. Diversity gain is another important metric which has been calculated. The achieved value of DG is very close to the ideal value (10 dB) over the two frequency bands. The simulated and measured radiation patterns were presented at three different frequencies and they showed a nearly stable omnidirectional behaviour.

These achievements indicate that the proposed antenna can be a good candidate to work within WLAN frequency band applications. Some differences have been noticed between the simulated and measured results which related to manufacturing and measurement tolerances.

138 References

[1] G. J. Foschini and M. J. Gans, “On Limits of Wireless Communications in a Fading

Environment when Using Multiple Antennas,” Wireless Personal Communications,

vol. 6, pp. 311–335, 1998.

[2] J. Thaysen and K. B. Jakobsen, “Wireless System Design considerations for low

antenna correlation and mutual coupling reduction in multi antenna terminals,”

European transactions on telecommunications, vol. 18, pp. 319-326, 2007.

[3] S. Papantonis and E. Episkopou, “Compact dual-band printed 2.5-shaped monopole

antenna for WLAN applications,” Progress in Electromagnetics Research C, vol. 24,

pp. 57-68, 2011.

[4] Y. Ban, Z. Chen, Z. Chen, K. Kang and J. L. Li, “Decoupled Hepta-Band Antenna

Array for WWAN/LTE Smartphone Applications,” in IEEE Antennas and Wireless

Propagation Letters, vol. 13, pp. 999-1002, 2014.

[5] M. R. Sharawi, “Printed MIMO antenna engineering,” Norwood: Artech House,

2014.

[6] A. C. K. Mak, C. R. Rowell and R. D. Murch, “Isolation Enhancement between Two

Closely Packed Antennas,” in IEEE Antennas and Wireless Propagation Letters, vol.

56, no. 11, pp. 3411-3419, Nov. 2008.

[7] Z. Li, Z. Du, M. Takahashi and K. Ito, “Reducing Mutual Coupling of MIMO

Antennas with Parasitic Elements for Mobile Terminals,” IEEE Antennas and

Wireless Propagation Letters, vol. 60, no. 2, pp. 473-481, Feb. 2012.

[8] K. S. Min, D. J. Kim and Y. M. Moon, “Improved MIMO Antenna by Mutual

Coupling Suppression between Elements,” The European Conference on Wireless

Technology, pp. 125-128, 2005.

139 [9] M. G. N. Alsath, M. Kanagasabai and B. Balasubramanian, “Implementation of

Slotted Meander-Line Resonators for Isolation Enhancement in Microstrip Patch

Antenna Arrays,” in IEEE Antennas and Wireless Propagation Letters, vol.12, pp.15-

18, 2013.

[10] S. Yoo and S. Kahng, “A Compact MIMO Antenna using ZOR Split Ring Resonator

Radiators with a Decoupling Structure,” Microwave Journal, pp. 1-11, Nov. 2011.

[11] M. S. Khan, M. F. Shafique, A. D. Capobiannco, E. Autizi and I. Shoaib, “Compact

UWB-MIMO Antenna Array with a Novel Decoupling Structure,” Proceedings of

2013 10th International Bhurban Conference on Applied Sciences and Technology

(IBCAST), pp. 347-350, Jan. 2013.

[12] I. Nadeem and D. Choi, “Study on Mutual Coupling Reduction Technique for MIMO

Antennas,” in IEEE Access, vol. 7, pp. 563-586, 2019.

[13] M. S. Sharawi, “Printed MIMO antenna systems: Performance metrics,

implementations and challenges,” Forum for Electromagnetic Research Methods

and Application Technologies (FERMAT), vol. 1, 2014 [Online]. Available:

http://www.e-fermat.org/

[14] J. Byun, J.-H. Jo, and B. Lee, “Compact dual-band diversity antenna for mobile

handset applications,” Microwave and Optical Technology Letters, vol. 50, pp. 2600-

2604, 2008.

[15] K. Chung and J. H. Yoon, “Integrated MIMO antenna with high isolation

characteristic,” Electronics Letters, vol. 43, pp. 199-201, 2007.

[16] S.-W. Su, C.-T. Lee, and F.-S. Chang, “Printed MIMO-antenna system using

neutralization-line technique for wireless USB-dongle applications,” in IEEE

Transaction on Antennas and Propagation, vol. 60, no. 2, pp. 456_463, Feb. 2012.

140 [17] S. Wang and Z. Du, “Decoupled Dual-Antenna System Using Crossed

Neutralization Lines for LTE/WWAN Smartphone Applications,” in IEEE Antennas

and Wireless Propagation Letters, vol. 14, pp. 523-526, 2015, doi:

10.1109/LAWP.2014.2371020.

[18] Y. Ou, X. Cai, and K. Qian, “Two-element compact antennas decoupled with a

simple neutralization line,” Progress in Electromagnetics Research, vol. 65, pp.

63_68, 2017.

[19] W. A. E. Ali and A. A. Ibrahim, “A compact double-sided MIMO antenna with an

improved isolation for UWB applications,” AEU-International Journal of

Electronics and Communications, vol. 82, pp. 7_13, Dec. 2017.

[20] H. Song et al., “Detectability of Breast Tumors in Excised Breast Tissues of Total

Mastectomy by IR-UWB-Radar-Based Breast Cancer Detector,” in IEEE

Transactions on Biomedical Engineering, vol. 66, no. 8, pp. 2296-2305, Aug. 2019.

[21] M. E. Yanik and M. Torlak, “Near-Field MIMO-SAR Millimeter-Wave Imaging

with Sparsely Sampled Aperture Data,” in IEEE Access, vol. 7, pp. 31801-31819,

2019.

[22] Y. Li, C. Sim, Y. Luo and G. Yang, “High-Isolation 3.5 GHz Eight-Antenna MIMO

Array Using Balanced Open-Slot Antenna Element for 5G Smartphones,” in IEEE

Transactions on Antennas and Propagation, vol. 67, no. 6, pp. 3820-3830, June

2019.

[23] R. G. Vaughan and J. B. Andersen, “Antenna diversity in mobile communications,”

in IEEE Transactions on Vehicular Technology, vol. VT-36, pp. 147–172, Nov.

1987.

141 [24] R. A. Bhatti, J. Choi and S. Park, “Quad-Band MIMO Antenna Array for Portable

Wireless Communications Terminals,” in IEEE Antennas and Wireless

Propagation Letters, vol. 8, pp. 129-132, 2009.

[25] S. Blanch, J. Romeu and I. Corbella, “Exact representation of antenna system

diversity performance from input parameter description,” in Electronics Letters,

vol. 39, no. 9, pp. 705-707, 1 May 2003.

[26] C. X. Mao and Q. X. Chu, “Compact Co-radiator UWB-MIMO Antenna with Dual

Polarization,” in IEEE Antennas and Wireless Propagation Letters, vol.62,

pp.4474 - 4480, Sept. 2014.

[27] H. Shin and J. H. Lee, “Capacity of multiple-antenna fading channels: Spatial fading

correlation, double scattering, and keyhole,” in IEEE Transactions on Information

Theory, vol. 49, pp. 2636-2647, Oct. 2003.

[28] S. Su, C. Lee and F. Chang, “Printed MIMO-Antenna System Using Neutralization-

Line Technique for Wireless USB-Dongle Applications,” in IEEE Transactions on

Antennas and Propagation, vol. 60, no. 2, pp. 456-463, Feb. 2012.

[29] A. Kumar et al., “High Isolation Compact Four-Port MIMO Antenna Loaded with

CSRR for Multiband Applications,” Frequenz 72, no. 9-10, pp. 415-427, 2018.

Retrieved 12 Aug. 2019, from doi: 10.1515/freq-2017-0276.

[30] E. Elkhazmi, C. H. See, N. A. Jan, R. A. Abd-Alhameed, N. Ali and N. J. McEwan,

“Design of a wideband printed MIMO monopole antenna using Neutralization lines

technique,” Asia-Pacific Microwave Conference, Sendai, Japan, 2014, pp. 983-

985.

142 [31] S. Zhang and G. F. Pedersen, “Mutual Coupling Reduction for UWB MIMO

Antennas with a Wideband Neutralization Line,” in IEEE Antennas and Wireless

Propagation Letters, vol. 15, pp. 166-169, 2016.

[32] J. Banerjee, R. Ghatak and A. Karmakar, “A compact planar UWB MIMO diversity

antenna with Hilbert fractal neutralization line for isolation improvement and dual

band notch characteristics,” 2018 Emerging Trends in Electronic Devices and

Computational Techniques (EDCT), Kolkata, 2018, pp. 1-6.

[33] M. N. Hasan, S Chu, S. Bashir, “A DGS monopole antenna loaded with U-shape

stub for UWB MIMO applications,” Microwave and Optical Technology Letters, pp.

1-9, 2019.

[34] S. R. Pasumarthi, B. K. Jagadeesh and P. A. Mallikarjuna, “Design of Tri-Band

MIMO Antenna with Improved Isolation using DGS and Vias,” Wireless Personal

Communications, vol. 110, no. 3, pp. 1523-1532, 2020.

143 CHAPTER FOUR

Compact Tri-Band MIMO Antenna with High Port Isolation for WLAN and WiMAX Applications using Neutralization Line and Defected Ground Structures

As mentioned in Chapter Three, the neutralization line technique can improve the isolation of the MIMO antenna by eliminating the coupling currents on the radiating element structures [1-3]. Consider a two-port MIMO antenna system in which port 1 is excited and port 2 is terminated. the induced current appearing at port two is caused by:

1- The close placement between the two ports which enables port 2 to receive an

appreciable value of the radiated field from port 1.

2- The surface current in the ground plane produced by port 1 is coupled to the

feedline of port 2 [4].

This current induction may raise the value of the correlation coefficient between the two ports. If the coupling is increased by this current induction then the isolation can be enhanced by cancelling them. A neutralization line is such a method to enhance the isolation by taking current at a specific point on the excited radiating element. This current should have almost the same magnitude and opposite phase to the total currents due to points 1 and 2 above. Nevertheless, the neutralization line technique has some certain limitations. The first one is the complication of selecting the attachment point of the NL on the radiator structure. The second one is the radiation pattern of the radiating element which could degrade the performance of this method. Additionally, the ground plane also has a significant effect on the radiation pattern, thus the overall isolation of the

MIMO antenna system could be improved by taking into account the currents on the

144 radiating structure of the radiators as well as the surface currents on the ground planes.

The radiating elements of the MIMO antenna may share the same ground plane [5] or at least they are connected at their ports [6]. In this design, a tri-band MIMO antenna with two identical monopole antennas was modelled and simulated to fulfil the requirements of WLAN and WiMAX applications. The proposed structure was excited using the coplanar waveguide (CPW) feeding technique and was printed on a low cost and easily available FR4 substrate. A new technique is presented to improve the isolation by integrating DGS and NL decoupling techniques in the same MIMO antenna structure where the NL is placed in the defected area to connect the ground planes. The antenna performance characteristics, in terms of scattering parameters, ECC, TARC, CCL, DG, peak gain, efficiency and antenna radiation pattern, were studied using simulation and measurement. Analysis of these characteristics indicated that the design is appropriate for

WLAN and WiMAX applications.

4.1 Antenna Design

The configuration of the proposed MIMO antenna is illustrated in Fig. 4.1. The design is made up of two identical monopole antennas. FR4 substrate with a thickness of 1.6 mm, dielectric constant (휖r) of 4.3 and loss tangent of 0.025 is used in this design. The total size of the proposed antenna is 47×36×1.6 mm3 which is suitable for mobile devices.

CPW lines are applied to feed the two radiating monopole antennas and they are cut from the upper corners to reduce proximity effects. The separation between the two monopoles is about 3 mm which is equivalent to 0.024 o at 2.45 GHz (where λo is the free space wavelength). In order to improve the isolation between the radiating elements, a new technique is applied by integrating DGS and NL decoupling techniques in the same

145 MIMO antenna structure, where the NL is placed in the defected area to connect the ground planes. The neutralization line has a total length of g/4 (where λg is the wavelength inside the substrate) at the isolation frequency (3.5 GHz). All the required dimensions of the proposed MIMO antenna are listed in Table 4.1.

Fig. 4.1. Schematic structure of the proposed MIMO antenna.

Table 4.2. Dimensions of the proposed MIMO antenna.

TableParameters 4.1. Dimen sionsValue of the proposedParameters MI MOValue antenn a.Parameters Value

L1 22mm w2 3mm wf 3.2mm L2 18.5mm w3 3mm Lg1 16mm L3 10.7mm w4 2mm Lg2 9mm L4 11.6mm Lf 17.5mm wg 8.9mm L5 7.5mm wf 3.2mm Ls1 4.5mm L6 5mm Lg1 16mm Ls2 2mm L7 7.5mm Lg2 9mm Ls3 1mm w1 3.25mm wg 8.9mm ws 0.5mm

146 4.2 Evaluation of S-Parameters of the Proposed MIMO

Antenna

The isolation performance and impedance bandwidth at various design stages are presented in this section. The evaluation starts by studying the effect of cutting a rectangular area from the middle part of the ground plane as shown in Fig. 4.2. The reflection coefficient characteristics for various values of the length (R) of the cut rectangle are shown in Fig. 4.3. It can be clearly seen that increasing the length of the defected area (R) shifts the resonant frequency at the three different bands towards the required resonant frequencies and it reaches 2.4, 3.5 and 5.5 GHz at R = 9 mm. The isolation of the second band is greatly improved by increasing the length of the defected area and it exceeds 15 dB at 5.2 GHz and 5.8 GHz, while it reaches 20 dB around 5.5

GHz, as shown in Fig. 4.4.

Fig. 4.2. Configurations of the proposed MIMO antenna showing the decoupling approaches. (a) With DGS, (b) With DGS and NL.

147

Fig. 4.3. Simulated reflection coefficients of the proposed MIMO antenna with different values of R (mm).

Fig. 4.4. Simulated transmission coefficients of the proposed MIMO antenna with different values of R (mm).

148 The next step is to evaluate the effect of the NL on the antenna performance. The simulated S-parameters of the proposed MIMO antenna with NL are plotted in Fig. 4.5 and Fig. 4.6. It can be noted that the bandwidth at the three different intended bands is expanded by inserting the neutralization line between the radiating elements of the antenna structures, as shown in Fig. 4.5. On the other hand, the mutual coupling is minimized considerably by adding the NL between the radiating elements at the three intended bands, as illustrated in Fig. 4.6.

Fig. 4.5. Simulated reflection coefficients of the proposed MIMO antenna with and without decoupling approaches.

Fig. 4.6. Simulated transmission coefficients of the proposed MIMO antenna with and without decoupling approaches.

149 4.3 Evaluation of Antenna Gain and Efficiency

To further show the performance of the proposed decoupling approaches, the antenna efficiency and gain of the proposed tri-band MIMO antenna were simulated using the

CST software package. The simulated gains of the antenna with and without NL are shown in Fig. 4.7. The results show the general trend of increased gain with frequency. It can be noticed that the antenna has achieved a 1 dB gain at the first band, 2 dB at the second band and 5 dB in the third band. The antenna efficiency for the proposed antenna with and without NL was also simulated, as shown in Fig. 4.8. It can be seen that there is an improvement in antenna efficiency in the second and the third bands by adding NL, to reach 90%. At the same time, there is decay by 6% in the first band. Otherwise, there is a dip in antenna gain and efficiency at 2.8 GHz. This is due to the high mismatch between antenna ports at this frequency, as shown in Fig. 4.5.

Fig. 4.7. Variation of calculated peak gain versus frequency for the proposed MIMO antenna with and without NL.

150

Fig. 4.8. Variation of calculated efficiency versus frequency for the proposed MIMO antenna with and without NL.

4.4 Results and Discussion

To verify the proposed decoupling method, a prototype has been fabricated using the same substrate and dimensions used in the simulation, as shown in Fig. 4.9. The compactness of the ultimate antenna is clearly observed. The S-parameter measurements were performed by using a vector network analyser whilst the radiation characteristics were tested inside the anechoic chamber. Comparisons between the simulated and measured results explored in the following sub-sections.

4.4.1 S-Parameters of MIMO Antenna With and Without NL

The simulated and the measured S-parameters of the antenna with and without the neutralization line were compared to validate the effectiveness of the neutralization line.

Fig. 4.10 shows a comparison between the simulation and measurement results for the

151 variation of S11 versus frequency. The experimental data shows good agreement with simulation results, with impedance bandwidths (|S11|≤ -10 dB) of 2.4–2.6, 3.3-4.4 and

4.8-6 GHz for the three achieved bands.

Fig. 4.9. The fabricated MIMO antenna design; (a) With NL, (b) Without NL.

(a)

152

(b)

Fig. 4.10. Simulated and measured reflection coefficients S11; (a) Without NL, (b) With

NL.

The simulated and measured transmission coefficients with the neutralization line are explained in Fig. 4.11. The measured coupling is near -18 dB at 2.4 GHz, -20 dB at 3.5

GHz and -25 dB at 5.5 GHz. Reasonable agreement between the results is achieved with some tolerance that can be attributed to reflections from the SMA connector and the objects near the antenna during the tests. To obtain better insight into the effectiveness of the neutralization line, the surface current distribution across the radiating elements of the proposed antenna were found using the CST software. Fig. 4.12 illustrates the obtained current distributions with and without neutralization line. These simulation results were calculated by exciting port 1 while port 2 was terminated with a matched load. It is clearly seen that the induced current in the feeding line at port 2 is high when the neutralization line is absent. On the other hand, the existence of the neutralization line will add a new

153 current path which gives an additional coupling to reduce the original coupling, as shown in Fig. 4.12-b.

Fig. 4.11. Simulated and measured transmission coefficients S21 for the proposed MIMO antenna with a NL.

(a) (b)

Fig. 4.12. Surface current distributions of the proposed antenna at 3.5 GHz; (a) Without NL;

(b) With NL.

154 4.4.2 Diversity Performance of the Proposed Antenna

The diversity performance of the proposed antenna with NL was examined in terms of scattering parameters, envelope correlation coefficient, total active reflection coefficient, channel capacity loss, diversity gain and 2D and 3D antenna radiation patterns. The usefulness of these metrics was discussed in detail in Chapter Two. ECC is linked to the amount of correlation between the radiating elements: it should have values as low as possible. Fig. 4.13 shows the measured and simulated results using the equation based on

S-parameters (3.1). It can be seen that a good agreement between the simulated and measured results was achieved and the value of ECC does not exceed 0.002 over the three frequency bands, confirming a good diversity performance.

The channel capacity loss of the proposed antenna with neutralization line was calculated, based on equation (2.22). The simulated and measured results were plotted in Fig. 4.14.

It can be easily seen that the CCL is less than 1 bit/s/Hz for the three frequency bands and there is a fair agreement between the simulated and measured results. These consequences signify that in rich scattering environments, the proposed MIMO antenna can transfer more data without increasing the input power, due to the low loss in the channel capacity.

Generally, the total active reflection coefficient can be used to indicate the importance of the stability in the resonant frequency, even when the phase difference between antenna elements is changed. TARC was evaluated using equation (3.2). During TARC evaluation, the input signal phase of port one was kept constant and port two changed from 0o to 180o with phase steps of 30o. Fig. 4.15 represents the variation of the TARC versus frequency, while the average value of TARC versus frequency is shown in Fig.

155 4.16. It can be seen that the three bands have TARC values less than -10 dB, which signifies good radiation performance.

Diversity Gain (DG) is another factor that is used to characterize the performance of a

MIMO antenna system. This metric can be defined as the amount of improvement in the diversity antenna system compared to a single antenna system in one diversity channel

[5]. Fig. 4.17 illustrates the simulated diversity gain of the proposed MIMO antenna. It can be clearly observed that the values of the DG for the three different bands are near to the ideal value of DG: this is equal to 10 dB.

Fig. 4.13. Simulated and measured ECC of the proposed MIMO antenna with NL.

156

Fig. 4.14. Simulated and measured CCL of the proposed MIMO antenna with NL.

Fig. 4.15. Calculated TARC with varying phases for the proposed MIMO antenna with

NL.

157

Fig. 4.16. The calculated average value of TARC for the proposed MIMO antenna with

NL.

Fig. 4.17. The simulated diversity gain of the proposed MIMO antenna with NL.

158 Finally, the prototyped design of the proposed antenna was tested in terms of far-field radiation patterns at four different frequencies. These patterns were measured in the two planes of XZ and YZ, in the case of exciting port 1 and terminating port 2. The practical results are compared to the simulated ones and are plotted in Fig. 4.18. These results show that at the four frequencies, the far-field radiations achieve a fair approximation to omnidirectional patterns. For further evaluation of the volumetric radiation patterns, the three-dimensional variation of the radiated field for the proposed antenna was calculated for the same frequency bands: this is shown in Fig. 4.19. This figure gives more appreciation of the field shape as compared to that of the 2-D representations.

159

Fig. 4.18. Simulated and measured radiation patterns of the proposed antenna for two planes [(1 and 2) y-x plane and (3 and 4) z-x plane] at (a) 2.4 GHz, (b) 3.5 GHz, (c) 5.2

GHz and (d) 5.8 GHz. Port 1 is excited and port 2 is terminated. Solid curve: simulated results, dashed curve: measured results, “1 and 3” are co-polar components and “2 and

4” are cross-polar components.

160

Fig. 4.19. The simulated 3-D pattern of the proposed MIMO antenna at (a) 2.4 GHz (b)

3.5 GHz, (c) 5.2 and (d) 5.8 GHz.

4.5 Comparison with Published Works

Table 4.2 gives a summarised comparison of the proposed tri-band monopole MIMO antenna against other multi-band monopole antennas that have appeared in the literature.

The MIMO antennas are tabulated concerning various criteria such as the antenna centre

161 frequency, the isolation between the elements, the overall electrical size of the antenna, the separation distance between the radiating elements in terms of λo (where λo is the free space wavelength at the lower centre frequency), antenna gain, antenna efficiency and

ECC. This comparison is not comprehensive but it is fairly representative for the state of the art of this technology. It can be clearly seen that the proposed tri-band MIMO antenna design explored in this chapter has some outstanding characteristics because of the compact size, geometric simplicity, high isolation in all the three bands and good antenna peak gain and efficiency. However; the main benefit achieved here is the simplicity of the design process of the integrated decoupling approach which led to these outstanding results.

Table 4.3. Comparison with other works.

Centre Peak Antenna Isolation Separation Ref. Frequency Electrical Size Gain Efficiency ECC (dB) distance (GHz) (dBi) (%) 2.425 >15 0.18 65 0.16 0.2λ × 0.48λ [7] 3.55 >14 o o 0.094λ 3.6 85 0.015 ×0.013λ o 5.6 >15 o 2.74 84 0.001 2.4 22 1.36 70.2 0.24λo×0.4λo [8] 3.5 18 0.096λo 2.01 76.6 <0.1 5.8 30 ×0.006λo 4.58 82.3 2.45 19 0.4λo×0.21λo 3 [9] 0.098λo 5.175 22 ×0.006λo 4.37 2.345 14 0.38λo ×0.37λo [10] 3.65 18 0.031λo <0.05 5.825 16 ×0.006λo 2.5 18 1 63 This 0.39λo ×0.3λo 3.865 20 0.024λo 2 92 <0.002 work 5.4 25 ×0.013λo 5 90

4.6 Conclusions

A low–profile tri-band MIMO antenna operating in WLAN (2.4/5.2/5.8 GHz) and

WiMAX 3.5 GHz bands has been designed. This antenna employs two printed Uni-Planar

162 monopole antennas. A new technique has been applied to improve the isolation between the radiating elements by integrating DGS with NL in one MIMO antenna structure. NL has been introduced in the defected area to connect the ground planes. The main benefit of this technique is to reduce the coupling caused by the surface current in the ground plane. The antenna has achieved impedance bandwidths (|S11|≤-10 dB) from 2.4–2.6, 3.3-

4.4 and 4.8-6 GHz, with isolation near 18 dB at 2.4 GHz, 20 dB at 3.5 GHz and 25 dB at

5.5 GHz. The envelope correlation coefficient has also been evaluated and found to be less than 0.002 in the three bands. Channel capacity loss is another parameter which has been calculated and presented: the obtained value of CCL is less than 0.7 bps/s/Hz in the three required bands. The simulated TARC has also been derived, the three required bands showed a TARC value less than -10 dB.

Diversity gain is another important metric which has been calculated. The achieved value of DG is very close to the ideal value (10 dB) over the two frequency bands. The simulated and measured radiation patterns were presented at four different frequencies and they showed nearly a stable omnidirectional behaviour. These achievements indicate that the proposed antenna can be a good candidate to work within WLAN and WiMAX frequency band applications. Some differences have been noticed between the simulated and measured results: these related to manufacturing and measurement tolerances.

163 References

[1] H. See, R. A. Abd-Alhameed, Z. Z. Abidin, N. J. McEwan, and P. S.

[2] Excell, “Wideband printed MIMO/diversity monopole antenna for WiFi/WiMAX

applications,” in IEEE Transactions on Antennas and Propagation, vol. 60, no. 4,

April 2012, pp.2028-2035.

[3] A. Diallo, C. Luxey, P. Le Thuc, R. Staraj and G. Kossiavas, “Enhanced two-antenna

structures for universal mobile telecommunications system diversity terminals,” IET

Microwaves, Antennas& Propagation, vol.2, no.1, pp.93-101, 2008.

[4] Q. Luo, H. M. Salgado and J.R. Pereira, “Compact printed monopole antenna array

for dual-band WLAN application,” IEEE International Conference on Computer as

a Tool (EUROCON), 27-29 April 2011, Lisbon, pp.1-4.

[5] I. Nadeem and D. Choi, “Study on Mutual Coupling Reduction Technique for MIMO

Antennas,” in IEEE Access, vol. 7, pp. 563-586, 2019.

[6] M. R. Sharawi, “Printed MIMO antenna engineering,” Norwood: Artech House,

2014.

[7] H.S. Singh and M.K. Meshram, “Printed Monopole Diversity Antenna for USB

Dongle Applications,” Wireless Personal Communications, vol. 86, no. 2, pp.771-

787, 2016.

[8] C. Luo, et al., “A decoupling method between two tri-band antennas for

WLAN/WiMAX applications,” IEICE Electronics Express, pp. 14-20170354, 2017.

(DOI: 10.1587/elex.14.20170354).

[9] X. Wu, G. Han and C. Feng, “Dual Frequency MIMO Antenna with Neutralization

Line,” 2019 Cross Strait Quad-Regional Radio Science and Wireless Technology

164 Conference (CSQRWC), Taiyuan, China, 2019, pp. 1-2, doi:

10.1109/CSQRWC.2019.8799229.

[10] R. Liu, X. An, H. Zheng, M. Wang, Z. Gao and E. Li, “Neutralization Line

Decoupling Tri-Band Multiple-Input Multiple-Output Antenna Design,” in IEEE

Access, vol. 8, pp. 27018-27026, 2020. doi: 10.1109/ACCESS.2020.2971038.

165 CHAPTER FIVE

Dual-Band Printed MIMO Antenna Decoupling Based on Defected Ground Structures and Parasitic Elements Techniques

Recently, the rapid expansion in the use of wireless communication systems, driven by the unprecedented evolution of their applications, has increased the demand for larger channel capacity with higher data rates. To fulfil these requirements, a promising solution of using MIMO antenna technology, which combines multiple antenna elements on the transmitting and receiving ends, has been applied. The main drawback associated with

MIMO antennas in low-profile handheld devices is the coupling between the radiating elements. On other words, it is crucial to take into consideration the mutual coupling/isolation parameter when two or more antenna elements are forced to be installed within the confined space that is available in a typical standard commercial enclosure [1]. Many approaches have been developed to improve the isolation factor in

MIMO antennas using various types of decoupling structures, as explained in Chapter

Two.

In this chapter, a low-profile dual-band MIMO antenna with enhanced isolation is proposed for WLAN applications. The designed antenna basically consists of two double concentric rectangular rings to achieve multi-band operation. A new method of combining two decoupling approaches within the antenna structure is used to minimize the coupling between the radiating antennas. The DGS approach is applied by creating a

166 defected areas in the corners of the ground plane to reduce the coupling by up to -20 dB at the centre frequency of the upper band. Additionally, five parasitic elements are introduced at the centre of the ground plane to act as an isolator and facilitate low mutual coupling for the lower frequency bands of the antenna.

5.1 MIMO Antenna Design

The configuration of the proposed 22 MIMO antenna is shown in Fig. 5.1. It consists of two double concentric rectangular rings to exhibit the self-similarity property where the final geometry of a certain shape can be formed by scaling down and repeating it a number of times. This property is used in this design to achieve multi-band operation, where the lower frequency band is controlled by the outer ring and the higher frequency band is influenced by the inner ring. The presented MIMO antenna is designed to operate at the frequency bands of WLAN (2.4-2.48 GHz, 5.15-5.35 GHz and 5.725-5.825 GHz). The two radiating elements with the feed lines are etched on the same surface of the substrate.

The characteristics of the substrate are 1.6 mm thickness, 4.3 relative permittivity and

0.025 loss tangent. The overall size of the proposed MIMO antenna is 44541.6 mm3.

The common ground plane is cut into three different parts by inserting two slots in between so that each radiating element could only couple to one part instead of sharing the same ground plane. This idea is used to reduce the reflection coefficient at the first band [2]. The DGS decoupling approach is applied by creating defected areas in the corners of the ground plane to reduce the coupling as well as to enlarge the bandwidth at the upper frequency band. Furthermore, five parasitic elements are added at the centre of the ground plane, two of them are connected to the middle part of the ground plane, as explained in Fig. 5.1-b, to act as an isolator and impose low mutual coupling for the lower

167 frequency band of the proposed MIMO antenna. All the required dimensions are listed in

Table 5.1.

Fig. 5.1. Configurations of the proposed antenna; (a) Front view; (b) Back view.

Table 5.1. The dimensions of the proposed MIMO antenna (mm).

Parameter Lf wf L1 L2 L3 L4 w1 w2 g

Dimension 19 2.4 21 16 9 10 2 1.5 1

Parameter Lg wg L w Lp Lpc wp

Dimension 18 16 15.5 13 19 18 1

5.2 Parametric Study

In order to clarify the effects and obtain the optimized values of the proposed design, parametric studies depending on the dimensions of the defected areas and slots are reported in this section.

168 5.2.1 S-Parameters of the Proposed MIMO Antenna with DGS

The investigations start with assessment of the effect of defecting the ground plane by cutting four different areas in the corners of the upper part of the ground plane as shown in Fig. 5.2. The S-parameter characteristics for various values of the height of the cut area

(U) are shown in Fig. 5.3 and Fig 5.4 respectively. It can be clearly seen that increasing the height of the cut area (U) has little effect on the reflection coefficient at the lower frequency band, as exhibited by the slight extension in the frequency bandwidth, while it increases the bandwidth of the reflection coefficient significantly at the second band.

Furthermore, the case for U = 4 mm produces the best results among the investigated ones, as shown in Fig. 5.3. On the other hand, the isolation between the antenna elements has significantly improved by increasing the value of the parameter U at the two intended bands and it reaches -20 dB at around 5.5 GHz, as shown in Fig. 5.4.

Fig. 5.2. Back view of the proposed MIMO antenna showing the four defected areas.

169

Fig. 5.3. Simulated reflection coefficients of the proposed MIMO antenna with different values of U (mm).

Fig. 5.4. Simulated transmission coefficients of the proposed MIMO antenna with different values of U (mm).

170 5.2.2 S-Parameters of the Proposed MIMO Antenna with

Different Slot Heights

The assessment starts with an investigation of the influence of etching two slots from the ground plane as shown in Fig. 5.5. The S-parameter characteristics for various values of the height of the cut slots (LS) are shown in Fig. 5.6 and Fig. 5.7 respectively. It can be seen that the cuts have little effect on the upper frequency band and it shows good correspondence with various values of LS while increasing the value of LS will lead to enhance the impedance bandwidth of the first band. The case for LS = 14 mm gives the best results among the investigated ones, as illustrated in Fig. 5.6. On the other hand, increasing the height of the slot (LS) has no significant effect on the isolation of the first and second bands, while the isolation between the two bands is reduced by increasing the height of the slot, as demonstrated in Fig. 5.7.

Fig. 5.5. Back view of the proposed MIMO antenna showing the two slots.

171

Fig. 5.6. Simulated reflection coefficients of the proposed MIMO antenna with different values of LS (mm).

Fig. 5.7. Simulated transmission coefficients of the proposed MIMO antenna with different values of LS (mm).

172 5.3 Parasitic Elements

Another approach to improve the isolation between the antennas in the MIMO system is to insert a parasitic element between them. This method will lead to attenuation of the mutual coupling between the antennas by introducing another coupling field which has the same magnitudes and opposite phase to the original one [1].

The design of the parasitic elements may be changed to achieve the wanted frequency band of coupling, operating bandwidth of the antenna and the amount of isolation improvement. There are two methods for designing the parasitic elements. The first method (such as those reported in [3-5]) is used for simple types of parasitic elements, for example, shorted or non-shorted stubs. On the other hand, some researchers, such as those in [6-8], used resonator types to design the parasitic element.

The simplified model that describes the principle of using parasitic elements to reduce mutual coupling in MIMO antenna systems is explained in Fig. 5.8 [4]. This figure shows two radiating elements placed in a compact area. Element A is excited by a current of Io and element B is terminated. Due to the small separation between the elements, a coupling current of αIo will excite element B, where α (which depends on the distance between the elements and the type of coupling between them) represents the coupling factor. By adding two parasitic elements between the radiating elements, two coupling paths will appear at element B. The first one is the original coupling path and the second one is a double coupling path (where the original current Io is coupled twice by parasitic elements 1 and 2 respectively). The overall coupling factor for the two parasitic elements is represented by β. Thus, the average value of the coupling current is expressed by βIo.

If the position of the two parasitic elements is the midpoint between element A and B, a

173 2 current of β Io will couple to element B. This means that the overall coupling on element

B is represented by [4]:

2 퐼푡표푡푎푙 = (훼 + 훽 )퐼표 (5.1)

The elimination of the mutual coupling can be achieved by changing the coupling factors

(α and β). This can be done by choosing the suitable antenna configuration and proper parasitic elements [1].

Fig. 5.8. Mutual coupling reduction mechanism through two parasitic elements: (a) Two radiating elements; (b) Two radiating elements with parasitic elements [4].

5.3.1 Parametric Study of the Proposed Antenna with Parasitic

Elements

To clarify the influences and obtain the optimized values of the proposed MIMO antenna design, a parametric study is carried out by changing the effective dimensions of the parasitic elements which are illustrated in Fig 5.1.

174 5.3.1.1 S-Parameters of the Proposed Antenna with Variations of Lp

The first parameter that is optimized to validate the effectiveness of the parasitic element is the length of the parasitic elements (Lp). The simulated S-parameters of the antenna with different dimensions of Lp are illustrated in Fig. 5.9 and Fig. 5.10 respectively. It can be clearly seen that when Lp =19 mm, the width of the second band increases and the mutual coupling also decreases to -17 dB at the first band and -25 dB at the second band.

Fig. 5.9. Simulated reflection coefficients with different values of Lp.

Fig. 5.10. Simulated transmission coefficients with different values of Lp.

175 5.3.1.2 S-Parameters of the Proposed Antenna with Different Variations of Wp

Another parameter has been investigated which is the width of the parasitic elements

(Wp). The simulated S-parameters of the antenna with different dimensions of Wp are illustrated in Fig. 5.11 and Fig. 5.12 respectively. It can be distinctly seen that the extension of the width of the five arms of the parasitic element will lead to a decrease in the mutual coupling at the first band and there is no significant effect on the second band.

Otherwise, the dips of the two bands of the reflection coefficients will change by extending the width of the arms of the parasitic element. For the previous reasons, taking

Wp=1 will balance the mutual couplings and reflection coefficients.

5.3.1.3 S-Parameters of the Proposed Antenna with Different

Variations of W

Examining the effect of the width of the middle part of the ground plane (W) is the third investigation factor which can help to achieve the best performance for the antenna. Fig.

5.13 and Fig. 5.14 illustrate the simulated S-parameters of the antenna with different dimensions of W. It can be noticed from Fig. 5.13 that the second band is shifted to the right side by increasing the width (W). On the other hand, the coupling between the antennas is increased by expanding the middle area of the ground plane. The optimum value can be chosen by making a trade-off between dimensions: this occurs at W=13 mm.

176

Fig. 5.11. Simulated reflection coefficients with different values of Wp.

Fig. 5.12. Simulated transmission coefficients with different values of Wp.

177

Fig. 5.13. Simulated reflection coefficients with different values of W.

Fig. 5.14. Simulated transmission coefficients with different values of W.

178 5.3.1.4 S-Parameters of the Proposed Antenna with and without

Parasitic Elements

Generally, the best way to validate the influence of the parasitic element in a MIMO antenna system is by comparing the simulated S-parameters of the antenna with and without it. Fig. 5.15 shows the simulated results of the reflection coefficients S11. It can be clearly seen that the first band will appear while inserting the parasitic element with

S11 less than -10 dB and also the two frequency bands were reduced in both directions.

On the other hand, the mutual coupling decreases in the two bands with the existence of the parasitic element and it reaches to -18 dB around the required bands, as illustrated in

Fig. 5.16.

Fig. 5.15. Simulated reflection coefficients S11 with and without parasitic element.

179

Fig. 5.16. Simulated transmission coefficients S21 with and without parasitic element.

5.4 Evaluation of Antenna Efficiency and Gain

The performance of the proposed MIMO antenna is further investigated through two useful measures, antenna efficiency and gain. They are calculated using the CST package in the case of the existence and the absence of the parasitic elements. The obtained results are plotted and compared in Fig. 5.17 and Fig. 5.18 respectively.

Fig. 5.17 shows that the radiation efficiency of the decoupled MIMO antenna is slightly higher than the coupled antenna by 3% and it achieves radiation efficiency equal to 83% at a frequency of 2.45 GHz. The value is quite high due to the good reduction in the reflection coefficient (down to about -20 dB) at this frequency. On the other hand, the proposed antenna with decoupling technique achieves radiation efficiency more than 80% around the centre frequency of the second band and with 2% improvement at 5.2 GHz compared to the coupled antenna case. 180 The peak gains of the proposed antenna with and without parasitic elements are illustrated in Fig. 5.18. It can be clearly seen that the peak gain increases significantly at the centre frequency of the two intended bands by adding the parasitic elements to the antenna structure, where the peak gain achieves 4.2 dB at the first centre frequency with an increase by 1.7 dB, and 4.1 dB at the second centre frequency: an increase by 0.8 dB.

Fig. 5.17. Variation of calculated efficiency versus frequency for MIMO antenna with and without parasitic elements.

Fig. 5.18. Simulated peak gain for the MIMO antenna with and without parasitic elements.

181 5.5 The Effect of the Ground Connection

In this section, the ground plane of the proposed MIMO antenna is extended to evaluate its impact on the S-parameter values. The extension has been achieved by adding a common ground plane to connect the two radiating elements with the middle part of the ground plane as shown in Fig. 5.19. The length (d) of the extended area is investigated to validate the effectiveness of extending the ground plane on the S-parameter values. The simulated S-parameters of the proposed MIMO antenna with different values of d are illustrated in Fig. 5.20 and Fig. 5.21 respectively. As can be seen, there is no effect of increasing the length of the extended ground on the first frequency band whilst the second band is slightly changed by increasing d.

Fig. 5.19. Back view of the proposed antenna showing the two slots.

Fig. 5.20. Simulated reflection coefficients with different values of d (cm).

182

Fig. 5.21. Simulated transmission coefficients with different values of d (cm).

5.6 Hardware Realization and Comparison with Simulations

The hardware verifications of simulated results for the proposed MIMO antenna are presented here. The designed antenna was prototyped using the same substrate and dimensions used in the simulation as shown in Fig. 5.22. The comparisons between the simulated and the measured results in terms of S-parameters are plotted in Fig. 5.23 and

Fig. 5.24 respectively. It can be clearly seen that there is a reasonable agreement between the results, allowing for some tolerance that can be attributed to reflections from the SMA connector and the objects near the antenna during the tests. The impedance bandwidth of the antenna covers the operating frequency bands from 2.15 to 2.52 GHz and from 4.5 to

6.45 GHz for reflection coefficient |S11| < -10 dB. These achieved bandwidths fulfil the requirements of the WLAN applications. On the other hand, the results of S21 for the two bands are better than -15 dB and it reaches to -25 dB around 2.4 GHz and -35 dB around

5.5 GHz.

183

Fig. 5.22. The fabricated MIMO antenna design with parasitic elements (a) Top view, (b)

Back view.

Fig. 5.23. Simulated and measured S11 of the proposed MIMO antenna with parasitic elements.

184

Fig. 5.24. Simulated and measured S21 of the proposed MIMO antenna with parasitic elements.

As explained in the previous chapters, the envelope correlation coefficient between the antenna radiating elements is another significant parameter in MIMO and diversity systems. This metric is directly connected with the spectral efficiency loss and performance degradation of diversity systems [9-11]. In general, two different ways can be used to compute the correlation coefficient of the MIMO antenna. The first one is by using the S-parameters of the antenna [12, 13] and the second one depends on the far- field radiation pattern of the antenna [4]. Due to the difficulties of computing and measuring the 3D far-field of the MIMO antenna, the first method of using S-parameters to evaluate the correlation coefficient of the two antennas is adopted. Depending on [12], the ECC (ρe) of a 2×2 MIMO antenna system could be calculated using equation (3.1).

The simulated and measured envelope correlation coefficients of the proposed MIMO antenna with parasitic elements are illustrated in Fig. 5.25, where it can be obviously seen that the value of ECC does not exceed 0.015 over the two frequency bands.

185

Fig. 5.25. Simulated and measured ECC of the proposed MIMO antenna with parasitic elements.

As explained in Chapter Two, the main benefit of a MIMO antenna system is that it produces an enhanced channel capacity compared to a SISO system in a multipath environment. The channel capacity of the MIMO system can be affected by the presence of uncorrelated Rayleigh fading: this will induce loss in channel capacity. On the other hand, the channel capacity can be improved by increasing the number of antennas inside the MIMO system. In this design, the simulated and measured channel capacity loss are compared and plotted in Fig 5.26 depending on the equation (2.22). It can be easily seen that the CCL is less than 1 bit/s/Hz for the two frequency bands and there is fair agreement between the calculated and measured results. These results mean that good impedance matching with low mutual coupling between the two antenna elements leads to a low capacity loss in the MIMO channel.

186

Fig. 5.26. Simulated and measured CCL of the proposed MIMO antenna with parasitic elements.

The total active reflection coefficient is a paramount parameter to evaluate the usefulness of this antenna as it is used to display a single curve that contains all the information of the S-parameters by manipulating all such for an N-port network to properly characterize the bandwidth and efficiency of the system. The computed TARC of the present design is shown in Fig. 5.27 and Fig. 5.28 respectively. Fig. 5.27 is evaluated using equation

(3.2) to cover the phase range from 0o to 180o with phase steps of 30o. The average value of TARC is shown in Fig 5.28. It can be clearly seen that the two bands appeared with

TARC values less than -10 dB.

Diversity Gain is one of the main metrics that affects the performance of a MIMO antenna system. This metric can be defined as the amount of improvement in the diversity antenna

187 system compared to a single antenna system in one diversity channel [1]. The simulated

DG of the proposed MIMO antenna is presented in Fig. 5.29 and it can be clearly seen that it is constantly greater than, or equal to, 9.7 dB, which conforms to the minimum limit of the defined DG value, i.e. 9.539.

Fig. 5.27. Calculated TARC with different phases of the proposed MIMO antenna with parasitic elements.

Fig. 5.28. The calculated average value of TARC for the proposed MIMO antenna with parasitic elements.

188

Fig. 5.29. Simulated diversity gain of the proposed MIMO antenna.

Finally, the radiation characteristic of the proposed antenna is also analysed at two different WLAN bands. These patterns were measured in the two planes of XZ and YZ, in the case of exciting port 1 and terminating port 2. The simulated and practical results are shown in Fig. 5.30 and they explain that at the two frequency bands the far-field radiations achieve stable omnidirectional patterns. A small discrepancy between the simulation and the measurement results is due to the fabrication, soldering, and measurement tolerances. For further evaluation of the volumetric radiation patterns, the three-dimensional variation of the radiated field for the proposed antenna was calculated at three different frequencies and is shown in Fig. 5.31. This figure gives more appreciation of the field shape as compared to that of the 2-D representations.

189

Fig. 5.30. Simulated and measured radiation patterns of the proposed antenna for two planes [(1 and 2) y-x plane and (3 and 4) z-x plane] at (a) 2.4 GHz (b) 5.2 GHz and (c)

190

Fig. 5.31. The simulated 3-D pattern of the proposed MIMO antenna at (a) 2.4 GHz (b)

5.2 GHz and (c) 5.8 GHz.

5.7 Comparison with Published Works

In this section, the performance parameters of the proposed MIMO antenna are compared with several other monopole MIMO antenna designs that have recently been published.

These parameters include the centre frequency of the antenna, the isolation between the radiating elements, the overall electrical size of the antenna, the separation distance

191 between the radiating elements in terms of λo (where λo is the free space wavelength at the lower centre frequency), antenna peak gain, antenna efficiency and ECC. The summary of this comparison is listed in Table 5.2. It can be noticed that the dimensions of the proposed antenna are the smallest among the compared antennas, with comparable values to [14] which indicates good compactness. The isolation of the proposed antenna is shown to provide a good value compared to the distance between the radiating elements. Low-level correlation coefficients, compared with other designs, have also been obtained. The antenna has comparable efficiency to the other designs and much higher than [15].

Table 5.2 Comparison with other works.

Centre Peak Antenna Isolation Separation Ref. Frequency Electrical Size Gain Efficiency ECC (dB) Distance (GHz) (dBi) (%) 0.6λo×0.38λo×0.0 [4] 1.92 >20 0.173λo 80

05λo 0.38λo ×0.38 λo [14] 2.29 >10 0.107 λo 90 <0.16 ×0.011 λo 2.5 0.41 λo ×0.25 λo 0.26 36.37 [15] >20 0.175λo <0.05 5.4 ×0.013 λo 3.1 65.70 2.76 0.73 λo ×1.1 λo 3.4 73 [16] >15 0.24λo 0.0341 5.45 ×0.007 λo 4.5 74 0.91 λo ×0.98λo [17] 6.85 >15 0.48λo 2 93 <0.2 ×0.022 λo 1.4 λ ×1.4 λ [18] 10.5 >20 o o 4 <0.03 ×0.056 λo This 2.335 >18 0.34 λo×0.42 λo × 2.2 70 0.108λo <0.015 work 5.475 >15 0.012λo 3.8 82

5.8 Conclusions

A compact dual-band MIMO antenna operating at WLAN bands (2.4/5.2/5.8 GHz) has been designed. This antenna has radiating elements consisting of two concentric double square rings printed symmetrically on the same substrate. A new method has been

192 adopted to suppress the unwanted coupling by introducing DGSs with five parasitic elements on the bottom layer of the radiating antennas. This new method has helped the

MIMO antenna to achieve a bandwidths from 2.15 to 2.52 GHz and from 4.5 to 6.45 GHz for reflection coefficient |S11| < -10 dB. On the other hand, the results of S21 for these two frequency bands are better than -15 dB and it has reached to -25 dB around 2.4 GHz and

-35 dB around 5.5 GHz. The envelope correlation coefficient has also been evaluated and found to be less than 0.015 in the two bands. Channel capacity loss is another parameter which has been calculated and presented: the obtained value is less than 1 bps/s/Hz in the two required bands. The simulated average TARCs has also been derived. It has been observed that the two bands appeared with TARC values less than -10 dB. Diversity gain is another important metric which has been calculated: the achieved value is very close to the ideal value (10 dB) over the two frequency bands. The simulated and measured radiation patterns were presented at three different frequencies and they showed nearly a stable omnidirectional behaviour. These achievements indicate that the proposed antenna can be a good candidate to work within WLAN frequency band applications. Some differences have been noticed between the simulated and measured results which are related to manufacturing and measurement tolerances.

193 References

[1] M. R. Sharawi, “Printed MIMO antenna engineering,” Norwood: Artech House,

2014.

[2] S. Zhang, P. Zetterberg and S. He, “Printed MIMO antenna system of four closely-

spaced elements with large bandwidth and high isolation,” in Electronics Letters,

vol. 46, no. 15, pp. 1052-1053, 22 July 2010.

[3] A. C. K. Mak, C. R. Rowell and R. D. Murch, “Isolation Enhancement between Two

Closely Packed Antennas,” in IEEE Antennas and Wireless Propagation Letters,

vol.56, no.11, pp. 3411-3419, Nov. 2008.

[4] Z. Li, Z. Du, M. Takahashi and K. Ito, “Reducing Mutual Coupling of MIMO

Antennas with Parasitic Elements for Mobile Terminals,” in IEEE Antennas and

Wireless Propagation Letters, vol.60, no.2, pp. 473-481, Feb. 2012.

[5] K. S. Min, D. J. Kim and Y. M. Moon, “Improved MIMO Antenna by Mutual

Coupling Suppression between Elements,” The European Conference on Wireless

Technology, pp. 125-128, 2005.

[6] M. G. N. Alsath, M. Kanagasabai and B. Balasubramanian, “Implementation of

Slotted Meander-Line Resonators for Isolation Enhancement in Microstrip Patch

Antenna Arrays,” in IEEE Antennas and Wireless Propagation Letters, vol.12,

pp.15-18, 2013.

[7] S. Yoo and S. Kahng, “A Compact MIMO Antenna using ZOR Split Ring Resonator

Radiators with a Decoupling Structure,” Microwave Journal, pp. 1-11, Nov. 2011.

[8] M. S. Khan, M. F. Shafique, A. D. Capobiannco, E. Autizi and I. Shoaib, “Compact

UWB-MIMO Antenna Array with a Novel Decoupling Structure,” Proceedings of

194 2013 10th International Bhurban Conference on Applied Sciences and Technology

(IBCAST), pp. 347-350, Jan. 2013.

[9] M. S. Sharawi, “Printed MIMO antenna systems: Performance metrics,

implementations and challenges,” Forum for Electromagnetic Research Methods

and Application Technologies (FERMAT), vol. 1, 2014 [Online]. Available:

http://www.e-fermat.org/

[10] G. J. Foschini and M. J. Gans, “On Limits of Wireless Communications in a Fading

Environment when Using Multiple Antennas,” Wireless Personal Communications,

vol. 6, pp. 311–335, 1998.

[11] R. G. Vaughan and J. B. Andersen, “Antenna diversity in mobile communication,”

in IEEE Transactions on Vehicular Technology, vol. 36, pp. 149- 172, 1987.

[12] R. A. Bhatti, J.-H. Choi, and S.-O. Park, "Quad-Band MIMO Antenna Array for

Portable Wireless Communications Terminals,” in IEEE Antennas and Wireless

Propagation Letters, vol. 8, pp. 129-132, 2009.

[13] S. Blanch, J. Romeu, and I. Corbella, “Exact representation of antenna system

diversity performance from input parameter description," in Electronics Letters, vol.

39, pp. 705-707, 2003.

[14] L. Wang, G. Wang, and Q. Zhao, “Suppressing mutual coupling of MIMO antennas

with parasitic fragment-type elements,” 46th European Microwave Conference

(EuMC), pp. 1303-1306, 2016.

[15] W. Wu et al, “A compact multiband MIMO antenna for IEEE 802.11 a/b/g/n

applications,” Progress In Electromagnetics Research Letters, vol. 84, pp. 59–65,

2019.

195 [16] S. S. Jehangir and M. S. Sharawi, “A Miniaturized Dual UWB Quasi-Yagi based

MIMO Antenna System Using a Defected Ground Structure,” 2018 IEEE

International Symposium on Antennas and Propagation & USNC/URSI National

Radio Science Meeting, Boston, MA, 2018, pp. 399-400,

[17] F. Amin, R. Saleem, T. Shabbir, S. Rehman, M. Bilal, and M. F. Shafique, “A

compact quad-element UWB-MIMO antenna system with parasitic decoupling

mechanism,” Applied Sciences, vol. 9, no. 11, p. 2371, Jun. 2019.

[18] Z. Tang, J. Zhan, X. Wu, Z. Xi, and S. Wu, “Simple ultra-wider-bandwidth MIMO

antenna integrated by double decoupling branches and square-ring ground

structure,” Microwave and Optical Technology Letters, vol. 62, no. 3, pp. 1259–

1266, Mar. 2019.

196 CHAPTER SIX

Isolation Enhancement of Compact Dual-Band Printed MIMO Antenna Using an Integration of Antenna Placement and Orientation with Metamaterial Resonators Techniques

With unprecedented use of portable devices such as mobiles and tablets, the existing capacity and bandwidth of current wireless communication systems have become insufficient, in particular for the future mobile generations. Thus, the fifth-generation

(5G) mobile networks, providing radically improved data rates, have recently come into the market. 5G technology will provide 10 Gbps or higher data rates which enables it to support various modern applications such as streaming video, interactive gaming, automated driving, Internet of things (IoT), smart cities, virtual reality (VR) and artificial intelligence (AI) [1-3]. MIMO systems can be considered as the backbone of the 5G technology due to their unique features of achieving higher data rates without consuming extra power or bandwidth in the multipath environment. On the other hand, it is a challenge to design a portable device that can hold several antenna systems within the limited structure of the device to fulfil all the required applications [4, 5]. Therefore, high bandwidth with a simple structure are very essential requirements for the 5G MIMO antenna system because they allow a simultaneous functioning of numerous system services [6]. Additionally, the geometry of the intended MIMO antenna should be compact to be part of handheld cellular devices.

In this chapter, a compact dual-band MIMO antenna with two arc-shaped radiating elements has been modelled and simulated to fulfil the requirements of WLAN and Sub-6

GHz 5G applications. The proposed structure is excited using the coplanar waveguide

197 (CPW) feeding technique and is printed on a low cost and easily available FR4 substrate.

The antenna placement and orientation technique is used to enhance the isolation in the sub-6 GHz 5G band while metamaterial resonators in the form of split-ring resonators

(SRRs) are introduced to reduce the antenna coupling in the WLAN band. The antenna performance characteristics, in terms of scattering parameters, ECC, TARC, CCL, DG, peak gain, efficiency and antenna radiation pattern, are studied using simulation and measurement.

6.1 Single Antenna Design

In this design, a dual-band asymmetric arc-shaped antenna has been modelled and simulated to fulfil the requirements of WLAN and sub-6 GHz 5G applications. The simple approach to design an antenna with a dual-band operation is by creating two elements with different lengths inside the antenna structure so that each element will resonate at the desired frequency by matching its length with a quarter of the wavelength of the operating frequency [7]. For antenna structures like the arc, two distinct bandwidths that meet the requirements for the WLAN and sub-6 GHz 5G applications can be obtained by adjusting the feed line position along the arc. Thus the arc can be considered as divided into two parts, where each part of them corresponds to one of the two bands. The geometry of the proposed antenna is shown in Fig.6.1. The antenna consists of two elements (short arc and long arc) with lengths that correspond to about a quarter of the wavelength at 2.4 GHz and 3.6 GHz respectively. The length of the long and short arcs can be calculated from equation (6.1):

198 1 1 푐 1 (6.1) 퐿 = 푒 = 4 4 푓푟 푘

Where L is the total length of the short or long arc, c is the speed of light, λe is the effective

2 wavelength in the substrate, k is a correction factor, taken as 푘 = 휖푟푒푓푓 [8], and the approximated value of 휖푟푒푓푓 is given by:

(휖푟 + 1) (6.2) 휖 = 푟푒푓푓 2

Equation (6.2) is mostly used in monopole designs, while the following equation (6.3) is used for microstrip antennas and feed lines [9].

1 ― (6.3) 휖푟 + 1 휖푟 — 1 2 휖푟푒푓푓 = ℎ + (1 + 12 ) 2 2 푤

The total length of the long arc is set to be 23.25 mm, to work at 2.45 GHz, while the total length of the short arc is set to be 12 mm, to work at 3.6 GHz. The antenna was constructed on the FR-4 substrate with a relative permittivity of 4.3, a thickness of 1.6 mm, and a loss tangent of 0.025. For the above substrate and using equations (6.1) and

(6.2), the ratio of quarter wavelength λe1 in the substrate to the total length of the long arc is 0.826 for the first band, whose centre is at 2.45 GHz, while the ratio of quarter wavelength λe2 in the substrate to the total length of the short arc is 1.09 at the considered centre frequency of 3.6 GHz. These results indicate the relation between the lengths of the arcs and wavelength. Moreover, with such relations, the designer can choose the centres of the two bands at the design stage. The overall dimensions of the single antenna are 27461.6 mm3. The optimized dimensions of the proposed single antenna are listed in Table 6.1. The radiating elements are fed separately by CPW lines.

199

Fig. 6.1. Geometrical configurations of the proposed single antenna.

Table 6.1. The optimized dimensions of the proposed single antenna.

Parameters Value Parameter value

A 46mm w 3.2mm

B 27mm w1 2.5mm

C 27.9mm OPQ 12mm

D 15mm OPR 23.25mm

E 13.9mm

To investigate the effect of two arms of the arc shape, simulations were performed to study the effect of changing the length of the long arm (OPR) only and keeping other parameters at the same values. The obtained reflection coefficient results are shown in

Fig. 6.2, where it can be seen that the first frequency band has been shifted to 2.4 GHz by increasing the length of the long arm and at the same time there is a very slight shift in the second band. Moreover, the other factor which is the length of the small arm (OPQ)

200 is also investigated to determine its effect on the reflection coefficient of the proposed antenna. Fig. 6.3 shows the obtained S11 results with different values of the length of the small arm. It can be clearly seen that there is no effect on the first band by increasing the length of the small arm, while the second frequency band is shifted to 3.6 GHz and has a lower return loss value of -23 dB.

Fig. 6.2. Variation of simulated S11 versus frequency with different length of OPR.

Fig. 6.3. Variation of the simulated S11 versus frequency with different length of OPQ.

201 It is crucial to evaluate the antenna gain and efficiency of the proposed antenna. These parameters can give extra information about antenna performance. The peak gain and radiation efficiency were evaluated using the CST software package. Fig. 6.4 shows that the peak gains of the proposed antenna are 2.5 dBi and 3.9 dBi at 2.4 GHz and 3.6 GHz, respectively. On the other hand, the radiation efficiencies of the proposed antenna reach

83% and 60% at the centre frequency of the lower and upper bands respectively, as shown in Fig. 6.4.

Fig. 6.4. Variation of the calculated gain and efficiency for the single element antenna.

6.2 Two Elements MIMO Antenna Design

The next step was to embed the antenna design into a two-element MIMO antenna configuration. The placement of the two antennas was carried out on the same FR4 substrate. A new approach of combining two different decoupling techniques was applied to improve the isolation between the radiating elements. The isolation in the upper frequency band was enhanced by examining different antenna placements and

202 orientations while the isolation in the lower frequency band was improved by introducing metamaterial resonators in the form of split ring resonators (SRRs) between the radiating antennas.

6.2.1 Antenna Placement and Orientation

Generally, closely positioned radiating elements within the MIMO antenna structure will have low isolation between them due to the system level ground plane as well as the radiated fields. This will affect the system performance by degrading their efficiencies as well as their correlation coefficients and, as a result, degrading the channel capacity that can be achieved. The easiest way to improve the system performance is by placing the radiating elements far apart from each other within the MIMO antenna structure. On the other hand, the orientation of the radiating elements can influence the phase of the coupling currents as well as the polarization of the radiated fields. Adjacent radiating elements can be arranged in quadrature with each other (i.e. 90o) to reduce the ground and field couplings. This should be approached properly via a total consideration of the antenna system proposed, because various antenna types have various coupling mechanisms and, as a result, these techniques can give different results. A few researches in this area have appeared in [10-17].

In this design, the analysis was carried out by fixing the position of one antenna and placing the second antenna element in four different orientations, as shown in Fig. 6.5 and the observations were made on the S-parameter characteristics at each proposed configuration, as outlined in Fig. 6.6 and Fig. 6.7 respectively. In the first proposed

MIMO antenna configuration, both the antenna elements were placed next to each other 203 whilst in the second proposed configuration, the first and the second antenna elements were rotated about 180o. In this configuration, the separation distance between the two elements was decreased compared to the first configuration. The third configuration was obtained by placing the second antenna element in front of the first antenna element. In the last configuration, the position of the first element was fixed at port 1 and the second element was rotated about 180o. The observation of their corresponding S-parameters shows that the S11 with a value less than -10 dB is covering the two bands for the four different orientations with significant bandwidth improvement in the two operating bands for the fourth proposed configuration ranging from 1.925 to 2.7 GHz and from 3.2 to 4.35

GHz, as illustrated in Fig. 6.6. Moreover, the isolation between the antenna elements for the four different orientations in the first band is just between -12 to -5 dB while isolation in the second band is enhanced significantly in the fourth proposed configuration compared to the three other configurations and it is well below -15 dB.

Fig. 6.5. Four different orientations of the proposed MIMO antenna.

204

Fig. 6.6. Simulated reflection coefficients of the four different proposed antenna orientations.

Fig. 6.7. Simulated transmission coefficients of the four different proposed antenna orientations.

205 6.2.2 Metamaterials

Metamaterials can be defined as an artificial material with unique electromagnetic properties that are not found in nature. The most important properties in these materials are the negative permittivity (휖) and/or negative permeability (µ): in other words, a negative refractive index (n). These unnatural properties of metamaterials lead to reversal of the direction of propagation inside it. This means the propagation of the electromagnetic waves will follow the left-hand rule (LH) instead of the right hand (RH), as inside natural materials. For this reason, these materials are known as left-hand materials [18]. Fig. 6.8 shows the different types of media and how the electromagnetic waves behave inside them. It can be noticed that the first type is called epsilon negative

(ENG) and the examples for this medium are metals and semiconductors. The second type of medium is known as double-positive (DPS) such as dielectric materials. The third type is named as double negative and an example of such a medium is metamaterial.

Finally, some ferrites are called Mu-negative materials (MNG) because they have a negative permeability (µ) [19]. The features of the metamaterials can be used for coupling reduction between the antennas in MIMO systems [20]. One of the most important features is the bandgap in the frequency response. The bandgap could work as a band- reject filter that can reduce mutual coupling and enhance the performance of the antenna

[21]. Several structures were used for coupling reduction’ for example, capacitively- loaded-loops (CLL) for those in [22, 23] or split ring resonators (SRR) and complementary split ring resonator (CSRR) [24-26], due to their capability to produce strong filtering behaviour (band-reject filter). In addition to isolation, metamaterials are also used to reduce the antenna dimensions for handheld and mobile devices [27], i.e. to miniaturize antennas.

206 µ

I II DPS ENG ε>0, µ>0 ε<0, µ>0 Most dielectric Metals, materials semiconductors

III IV DNG MNG ε<0, µ<0 ε>0, µ<0 Unnatural Some ferrites materials

Fig. 6.8. Electromagnetic wave behaviour in different media [19].

6.2.2.1 Theoretical Aspect

Generally, the characteristic matrix of a stratified substrate is represented by a constant permittivity (휖), permeability (휇) and refractive index 푛 = 휖휇, Therefore, for a two-port network showing incident waves (a1, a2) and reflected waves (b1, b2) as explained in Fig.

6.9, the refractive index n can be calculated by equation (6.5) [28-30]:

Fig. 6.9. Simple two-port network.

207 푏1 푆11 푆12 푎1 (6.4) = [푏2 ] [푆21 푆22 ][푎2 ]

± 1 ―1 2 2 (cos 1 ― 푆11 + 푆 21 (6.5) 푛 = + 2휋푚) 푘0ℎ 2푆21

Where h represents the thickness of the substrate and ko is the free-space propagation constant.

The extraction of the constitutive parameters (휖 and 휇) of any material usually needs experimental tests or analytical models [31]. The Drude-Lorentz model, known as the dispersion model is very accurate [32], in which the magnetic permeability and electric permittivity are extracted analytically using a mathematical model [33]. In this chapter, the constitutive parameters of the unit cell of metamaterial are obtained using the well- developed characterization method of metamaterials known as the standard retrieval procedure [34], where the effective permittivity and effective permeability refractive values of the metamaterial unit cell can be extracted from the S-parameters assuming that this latter are symmetric with respect to the (x-y) plane. This means that S11 = S22 and S21

= S12. The relative impedance of the thin dielectric slab is calculated with respect to the

S-parameters using the following formula [30]:

2 2 (1 + 푆11) ― 푆 21 (6.6) 푧 =± 2 2 (1 ― 푆11) ― 푆 21

The ambiguity of the signs of the refractive index n and the impedance z then can be computed, subject to the following conditions:

208 푛" ≥ 0

푧’ ≥ 0 where (.)’ and (.)” represent the real and the imaginary part operators respectively.

Finally, the complex permittivity (휖) and permeability (휇) are given by [30]:

푛 (6.7) 휖 = 푧

휇 = 푛.푧 (6.8)

6.2.2.2 Metamaterial Unit Cell Specification and Design

Generally, the behaviour of the split ring resonator can be considered as an electric dipole which can be excited by external electric flux. In MIMO antenna applications, the presence of SRR with its feature of open rings between the radiating elements will lead to production of an effective negative permeability at the resonant frequency. For this reason, high isolation between the radiating elements can be achieved [35]. The schematic configuration of the SRR unit cell is shown in Fig. 6.10. The SRR unit cell was designed and simulated using CST Microwave Studio, depending on [36], while the Matlab program was used here to calculate constitutive parameters of the unit cell, such as effective permittivity and permeability, from the S-parameters. It can be clearly seen from

Fig. 6.10 that the unit cell in this chapter has symmetrical periodicity in the central section. This leads to generation of negative permittivity and permeability values at the same time, as shown in Fig. 6.12 and Fig. 6.13 respectively. On the other hands, a unit cell with no periodicity normally produces negative permeability only, as explained in

[37]. The unit cell simulation setup is explained in Fig. 6.11 and it shows that the SRR

209 structure is positioned in the middle of the waveguide. The front and back sides of a waveguide are defined as PEC (perfectly electrically conducting) walls, while its top and bottom sides are attributed as PMC (perfectly magnetically conducting) walls, and the other two sides are used for the signal excitation. The main application for the metamaterial resonator in this chapter is to increase the isolation between the radiating elements inside the MIMO antenna structure. Consequently, four unit cells of size 9×5.5 mm2 are inserted between the radiating elements of the proposed MIMO antenna. The optimized dimensions of the proposed unit cell are given in Table 6.2. The transmission and reflection coefficients of the metamaterial unit cell are shown in Fig. 6.14. It can be obviously seen that the metamaterial unit cell has a resonant frequency at 2.45 GHz.

Fig. 6.10. The geometrical configuration of SRR unit cell.

210

Fig. 6.11. The simulation setup of SRR unit cell.

Table 6.2. The optimized dimensions of the proposed SRR unit cell.

Parameters Value Parameter value

H 5.5mm Wp 4mm

F 9.75mm Wu 2.5mm

h1 1.5mm gu 0.5mm

h2 1.5mm s 0.5mm

Fig. 6.12. The calculated permittivity of the proposed SRR unit cell.

211

Fig. 6.13. Calculated permeability of the proposed SRR unit cell.

Fig. 6.14. Transmission and reflection coefficients of the proposed SRR unit cell.

6.2.2.3 Antenna Layout with Metamaterial Resonators

The next step in this design is to justify the effectiveness of metamaterial as a decoupling approach. The fourth proposed configuration was adopted to be implemented with metamaterial resonators. The overall size of the fourth proposed MIMO antenna is

54×46×1.6 mm3 while the separation distance between the two radiating elements is equal to 6.65 mm (0.0532 o at 2.4 GHz), where o is the free-space wavelength. In order to reduce the unwanted coupling between the radiating elements, four SRR metamaterial

212 unit cells are introduced between the two elements, as clarified in Fig. 6.15. The

optimized dimensions of the Parameters Value Parameter value proposed MIMO antenna are A 46mm F 17.2mm

B 54mm w 3.2mm given in Table 6.3.

C 27.9mm w1 2.5mm

D 15mm OPQ 12mm

E 13.9mm OPR 23.25mm

Fig. 6. 15. Configurations of the proposed antenna; (a) Antenna without metamaterial resonators (b) Antenna with metamaterial resonators.

Table 6.3. The optimized dimensions of the proposed MIMO antenna.

213 6.2.2.3 Parametric Study

To clarify the effects of inserting the SRRs between the radiating elements and obtain the optimum values of the proposed design, a parametric study was carried out in the following sub-sections, focusing on the number of SRR unit cells between the radiating elements of the proposed MIMO antenna.

6.2.2.3.1 S-Parameters of The Proposed Antenna with and without

SRRs

The effectiveness of the metamaterial resonators has been validated by comparing the simulated S-parameters of the antenna with and without such resonators. Fig. 6.16 and

Fig. 6.17 respectively show the simulated results of the transmission and reflection coefficients. It can be clearly observed in Fig. 6.16 that the two frequency bands still appear after inserting the unit cells and there is bandwidth extension in the two operating frequency bands. On the other hand, the mutual coupling has decreased in the first band with the existence of the metamaterial resonators and it reaches to -35 dB around 2.5 GHz while the dip of the second band has shifted toward the first band but it still achieves mutual coupling less than -15 dB around 3.4 and 3.6 GHz, as illustrated in Fig. 6.17.

214

Fig. 6.16. Simulated reflection coefficients with and without metamaterial resonators.

Fig. 6.17. Simulated transmission coefficients with and without metamaterial resonators.

215 6.2.2.3.1 S-Parameters of The Proposed MIMO Antenna With Three

Variations of Metamaterial Resonators

To validate the effectiveness of the number of the metamaterial resonators, the simulated

S-parameters of the antenna with different numbers of such resonators (SRRs) are illustrated in Fig. 6.18 and Fig. 6.19 respectively. Fig. 6.18 shows that increasing the number of the metamaterial resonators will lead to a decrease in the mutual coupling around 2.4 GHz, reaching -35 dB with four such resonators, while the two bands still appear with |S11| less than -10 dB, as shown in Fig. 6.19.

Fig. 6.18. Simulated reflection coefficients of the proposed MIMO antenna with a different number of metamaterial resonators.

216

Fig. 6.19. Simulated transmission coefficients of the proposed MIMO antenna with a different number of metamaterial resonators.

6.3 Evaluation of Antenna Efficiency and Gain

The antenna efficiency and gain are useful measures that may be used to describe the performance of the proposed antenna. There is a direct relation between each other. The efficiency factor takes both the conductor loss and dielectric loss into account, but not the impedance mismatch between the feed line and the antenna. On the other hand, the gain of the antenna takes into account the efficiency of the antenna as well as its directional capabilities.

6.3.1 Efficiency Calculations

It is interesting to evaluate the efficiency of the proposed MIMO antenna system after using decoupling techniques. Therefore, the radiation efficiency of the proposed MIMO antenna with decoupling was simulated and compared with two other radiation

217 efficiencies. The first radiation efficiency is for the single printed monopole antenna and the second radiation efficiency is for the two closely spaced printed MIMO antennas without decoupling. The proposed MIMO antenna with decoupling achieves an efficiency equal to 80% over the first band, a drop of 5% and 3% compared to the efficiencies of the single antenna and MIMO antenna without decoupling respectively.

This drop in the value of the efficiency could be related to the metamaterial resonator due to its negative value of permittivity and permeability at this frequency. On the other hand, the efficiency of the proposed MIMO antenna with decoupling is improved considerably over the second band and it reaches 73% at 3.6 GHz, with an improvement of 10% compared with the efficiencies of the single antenna and MIMO antenna without decoupling. The obtained efficiencies versus frequency are shown in Fig. 6.20.

Fig. 6.20. Variation of calculated efficiency versus frequency for the single antenna,

MIMO antenna without decoupling and MIMO antenna with decoupling.

218 6.3.2 Peak Gain Calculations

The peak gain of each of the proposed antennas was calculated from the far-field patterns using the CST package, and the obtained gains versus frequency are shown in Fig. 6.21.

The gain has the general trend of increasing with frequency except between the two bands. This increase is due to the fact that at higher frequencies the antenna size in terms of wavelength becomes larger. The peak gain of the proposed MIMO antenna with decoupling is about 4 dB at the first band with a notable difference to the values of the peak gains of the single antenna and MIMO antenna without decoupling. Moreover, the peak gain of the proposed MIMO antenna with decoupling increases at the second band and it achieves 4.5 dB.

Fig. 6.21. Simulated peak gain for the single antenna, MIMO antenna without decoupling and MIMO antenna with decoupling.

219 6.4 Equivalent Circuit Model

To give a profound comprehension of the global configuration, the equivalent circuit model is provided in Fig. 6.22. Three different structures are added together: the radiating element of the proposed antenna is modelled by two resonator structures to fulfil the dual- band property; the decoupling structure of the metamaterial unit cell is modelled as an

RLC circuit; and the coupling between the radiating elements at the two bands is represented by two RLC circuits. All the optimized parameters of the proposed circuit model are extracted from the CST simulation results and are given in Fig. 6.22. The comparisons between the circuit model and full-wave electromagnetic (EM) model in terms of S-parameters are shown in Fig. 6.23 and Fig. 6.24 respectively. Reasonable consistency is observed in the frequency region of interest between the circuit model and full-wave EM simulated results.

Fig. 6.22. Equivalent circuit model of the proposed MIMO antenna.

220

Fig. 6.23. Comparison of reflection coefficients (S11) for EM model and circuit model of the proposed MIMO antenna.

Fig. 6. 24. Comparison of transmission coefficients (S21) for EM model and circuit model of the proposed MIMO antenna.

221 6.4 Results and Discussions

To verify the simulated outcomes, the proposed MIMO antenna was fabricated and tested.

The prototype design of the proposed antenna is displayed in Fig. 6.25. The scattering parameters in terms of reflection coefficient (S11) and transmission coefficient (S21) were measured using a vector network analyser (VNA) while the radiation patterns were measured using the anechoic chamber. These measurement results were used to analyse the performance of the proposed antenna. The measured achievements in terms of scattering parameters, envelope correlation coefficient, total active reflection coefficient, channel capacity loss, radiation pattern and diversity gain are explained in the following paragraphs.

The comparison between the simulated and the measured results of the S-parameters have been illustrated in Fig. 6.26 and Fig. 6.27 respectively. These results show a fair agreement between them with some variance due to the measurement tolerances. The measured reflection coefficients with |S11| less than -10 dB are achieved from 2.15-2.75

GHz for the lower band and from 3.1 to 4.4 GHz for the upper band. On the other hand, the mutual coupling values are -30 dB at 2.4 GHz, -20 dB at 3.4 GHz and -19 dB at 3.6

GHz, as shown in Fig. 6.27.

222

(a)

(b)

Fig. 6.25. The prototype of the proposed antenna, (a) Front view; (b) Back view.

Fig. 6 26. Simulated and measured reflection coefficient S11 of the proposed antenna.

223

Fig. 6.27. Simulated and measured transmission coefficient S21 of the proposed antenna.

As explained in the previous chapters, the envelope correlation coefficient can be used to examine the extent to which the radiating elements are independent and they rely on their individual performances. This parameter is directly connected with the spectral efficiency loss and performance degradation of diversity systems [21]. The ECC is calculated based on the S-parameter equation (3.1). The simulated and measured envelope correlation coefficients of the proposed MIMO antenna with decoupling are plotted in Fig. 6.28. It can be clearly seen that the value of ECC does not exceed 0.05 over the two frequency bands, thus providing the high isolation required for simultaneous operation.

Basically, the channel capacity of a MIMO system could be improved by increasing the number of antennas. On the other hand, the existence of the uncorrelated Rayleigh fading may lead to a loss in the channel capacity. For two elements of MIMO systems, the channel capacity loss could be calculated by using the correlation matrix mentioned in

[38-40]. The simplified form of channel capacity loss was given in equation (2.22). The

224 computed and measured channel capacity loss of the proposed MIMO antenna is illustrated in Fig. 6.29. In this figure, the capacity loss is less than 0.5 bit/s/Hz for the two bands and there is a very good agreement between the calculated and measured results.

Fig. 6.28. Simulated and measured ECC of the proposed MIMO antenna.

Fig. 6.29. Simulated and measured capacity loss of the proposed MIMO antenna.

225 The total active reflection coefficient was evaluated as another defining parameter in the usefulness of this antenna: this is used to display a single curve that contains all the information of the S-parameters by manipulating the S-parameters for an N-port network.

The computed TARC of the present design is shown in Fig. 6.30. This figure is plotted using equation (3.2) to cover the phase range from 0o to 180o with phase steps of 30o. The average value of TARC is shown in Fig. 6.31. It can be clearly seen that the two operating frequency bands appear with TARC values less than -10 dB.

Diversity Gain can be defined as the amount of improvement in the diversity antenna system compared to SISO systems [41]. It is very important to calculate DG due to its effects on the performance of MIMO antenna systems. Fig. 6.32 represents the simulated

DG of the proposed MIMO antenna. It can be clearly seen that the DG values for the two frequency bands are very close to the ideal value of DG = 10 dB.

Fig. 6.30. The calculated values of TARC with different phases.

226

Fig. 6.31. The calculated average value of TARC.

Fig. 6.32. The simulated diversity gain of the proposed MIMO antenna.

Finally, the prototyped design of the proposed antenna was tested in terms of far-field radiation patterns at three different frequencies (2.4 GHz, 3.4 GHz and 3.6 GHz) by applying the same procedure of Chapter Three. The practical results are shown in Fig.

227 6.33 and they show that, at the three frequencies, the far-field radiations achieve stable omnidirectional patterns. For further evaluation of the volumetric radiation patterns, the three-dimensional variation of the radiated field for the proposed antenna was calculated at the three frequencies and is plotted in Fig. 6.34. This figure gives more appreciation of the field shape as compared to that of the 2-D representations.

Fig. 6.33. Simulated and measured radiation patterns of the proposed antenna for two planes [(1 and 2) y-x plane and (3 and 4) z-x plane] at (a) 2.4 GHz (b) 3.4 GHz and (c)

3.6 GHz. Port 1 is excited and port 2 is terminated. Solid curve: simulated results, dashed curve: measured results, “1 and 3” are co-polar components and “2 and 4” are cross-polar components.

228

Fig. 6.34. The simulated 3-D pattern of the proposed MIMO antenna at (a) 2.4 GHz (b)

3.4 GHz and (c) 3.6 GHz.

6.5 Comparison with Published Works

There are various designs for MIMO antennas which differ vastly in their performance characteristics and sizes in terms of the wavelength at the centre frequency. The characteristics of the proposed MIMO antenna are compared in Table 6.4 with those of other MIMO designs that have been published in the literature. All the antennas in this comparison are planar antennas to ensure a like for like comparison. This comparison depends on the centre frequency of the antenna, the isolation between the radiating elements, the overall electrical size of the antenna, the separation distance between the

229 radiating elements in terms of λo (where λo is the free space wavelength at the lower centre frequency), antenna peak gain, antenna efficiency and ECC. It can be clearly seen from

Table 6.4 that the proposed MIMO antenna has a competitive performance as regards the isolation and size. On the other hand, the antenna has comparable efficiency, ECC and gain to the other designs, although much higher than [22]. These features confirm that this design is a good candidate for modern communication devices.

Table 6.4. Comparison with other works.

Centre Peak Antenna Isolation Separation Ref. Frequency Electrical Size Gain Efficiency ECC (dB) distance (GHz) (dBi) (%) 0.835 10 0.139 λo×0.27 λo -2.8 35 [22] 0.06 λo <0.01 2.65 18.9 ×0.004 λo 5.5 67 0.31 λo×0.22 λo [42] 2.45 >15 0.06 λo 0.002 ×0.13 λo

0.69 λo×0.52 λo [43] 2.6 >25 0.043 λo 5 82 <0.02 ×0.006 λo 2.58 0.63 λo×0.4 λo 2.8 95 [44] 30 0.044 λo <0.01 5.27 ×0.013 λo 2.9 92

0.98 λo×0.35 λo [45] 5.35 >35 <4 <10-3 ×0.028 λo This 2.45 20 0.44 λo×0.37 λo 3.8 81 0.0532 λo <0.05 work 3.75 30 ×0.013 λo 4.9 73

6.6 Conclusions

In this work, a 2×2 dual-band MIMO antenna for WLAN, and sub-6 GHz 5G applications is presented. The antenna placement and orientation decoupling approach, alongside metamaterial-based isolation enhancement structures have been used in this design. In this design, applying the first decoupling method has helped to improve the isolation in the second band of operation whilst the second method has enhanced the isolation in the first band. The proposed MIMO antenna has been fabricated and tested and it has achieved bandwidths from 2.15-2.75 GHz for the lower band and from 3.1 to 4.4 GHz

230 for the upper band, with isolation values around 30 dB at 2.4 GHz, 20 dB at 3.4 GHz and

19 dB at 3.6 GHz. The envelope correlation coefficient has also been evaluated and found to be less than 0.05 in two bands. Channel capacity loss is another parameter which has been calculated and presented. The obtained value of CCL is less than 0.5 bps/s/Hz in the two required bands. The simulated average TARC has been also derived. It has been observed that the simulated result with TARC value less than or equal to -10 dB covers the frequency band from 2.32 GHz to 4.79 GHz. Diversity gain is another important metric which has been calculated: the achieved value is very close to the ideal value (10 dB) over the two frequency bands. The simulated and measured radiation patterns were presented at three different frequencies and they showed nearly a stable omnidirectional behaviour. These achievements indicate that the proposed antenna can be a good candidate to work within WLAN and sub-6 GHz 5G frequency band applications. Some differences have been noticed between the simulated and measured results which can be explained by manufacturing and measurement tolerances.

231 References

[1] H. C. Huang, “Overview of antenna designs and considerations in 5G cellular

phones,” in 2018 International Workshop on Antenna Technology (iWAT). IEEE,

2018, pp. 1-4.

[2] T. Sharma, K. Ritesh, N. Chauhan and S. Agarwal, “Analogous study of 4G and

5G,” 2016 3rd International Conference on Computing for Sustainable

Global Development (INDIACom), New Delhi, 2016, pp. 2137-2140.

[3] M. Shuhrawardy, M. H. Miah Chowdhury and R. Azim, “A Four-element Compact

Wideband MIMO Antenna for 5G Applications,” 2019 International Conference on

Electrical, Computer and Communication Engineering (ECCE), Cox'sBazar,

Bangladesh, 2019, pp. 1-5.

[4] H. Zou, Y. X. Li, C.-Y.-D. Sim, and G. L. Yang, “Design of 8×8 dual-band MIMO

antenna array for 5G smartphone applications,” International Journal of RF

Microwave Computer-Aided Engineering, vol. 28, no. e21420, Nov. 2018.

[5] J. Li et al., “Dual-Band Eight-Antenna Array Design for MIMO Applications in 5G

Mobile Terminals,” in IEEE Access, vol. 7, pp. 71636-71644, 2019.

[6] Y. Li, C. Wang, H. Yuan, N. Liu, H. Zhao and X. Li, “A 5G MIMO Antenna

Manufactured by 3-D Printing Method,” in IEEE Antennas and Wireless

Propagation Letters, vol. 16, pp. 657-660, 2017.

[7] S. Papantonis and E. Episkopou, “Compact dual-band printed 2.5-shaped monopole

antenna for WLAN applications,” Progress in Electromagnetics Research C, vol.

24, PP. 57-68, 2011.

232 [8] K. H. Sayidmarie, L. S. Yahya, “Design and Analysis of Dual Band Crescent Shape

Monopole Antenna for WLAN Applications,” International Journal of

Electromagnetics and Applications, vol. 3, no. 4, pp. 96-102, 2013.

[9] Kumer, and K. P. Ray, “Broadband microstrip antennas,” Artech House, 2003.

[10] M. K. T. Al-Nuaimi and W. G. Whittow, “Performance investigation of a dual

element IFA array at 3 GHz for MIMO terminals,” in 2011 Loughborough Antennas

& Propagation Conference. IEEE, Nov. 2011, pp. 1-5.

[11] J. Xiong, M. Zhao, H. Li, Z. Ying and B. Wang, “Collocated Electric and Magnetic

Dipoles With Extremely Low Correlation as a Reference Antenna for Polarization

Diversity MIMO Applications,” in IEEE Antennas and Wireless Propagation

Letters, vol. 11, pp. 423-426, 2012.

[12] S.Y. Lin and I.H. Liu, “Small inverted-U loop antenna for MIMO applications,”

Progress in Electromagnetics Research, 34, pp.69-84, 2013.

[13] G. T. Jeong, S. Choi, K. H. Lee, andW. S. Kim, “Low profile dual wideband MIMO

antenna with low ECC for LTE and Wi-Fi applications,” International Journal of

Antennas and Propagation, Hindawi Publishing Corporation, vol. 2014, pp. 1-6,

2014.

[14] M. U. Khan, M. S. Sharawi, “A dual band microstrip annular slot based MIMO

antenna system,” Microwave & Optical Technology Letters, vol. 57, no. 2, pp. 360-

364, Feb. 2015.

[15] L. Malviya, R. K. Panigrahi, and M. V. Kartikeyan, “A 2×2 dual-band MIMO

antenna with polarization diversity for wireless applications,” Progress In

Electromagnetic Research C, vol. 61, pp. 91–103, Jan. 2016.

233 [16] R. Mathur and D. Santanu, “Compact 4-Port MIMO/Diversity Antenna with Low

Correlation for UWB Application,” Frequenz, 2018.

[17] Aw, M. S., K. Ashwath, and Tanweer Ali. “A compact two element MIMO antenna

with improved isolation for wireless applications,” Journal of Instrumentation, vol.

14, no. 06, 2019.

[18]Y. Dong and T. Itoh, “Metamaterial-Based Antennas,” in Proceedings of the IEEE,

vol. 100, no. 7, pp. 2271-2285, July 2012.

[19] T. J. Cui; D. R. Smith and R. Liu, “Metamaterials, theory, design and Applications,”

ISBN: 978-1-4419-0572-7, Springer New York Dordrecht Heidelberg London 2010.

[20] K. M. K. H. Leong, C. J. Lee and T. Itoh, “Compact Metamaterial Based Antennas

for MIMO Applications,” 2007 International workshop on Antenna

Technology: Small and Smart Antennas Metamaterials and Applications,

Cambridge, 2007, pp. 87-90.

[21] M. R. Sharawi, “Printed MIMO antenna engineering,” Norwood: Artech House,

2014.

[22] M. S. Sharawi, A. B. Numan and D. N. Aloi, “Isolation Improvement in a Dual-

Element MIMO Antenna System Using Capacitively Loaded Loops,” Progress in

Electromagnetics Research, vol. 134, pp. 247-266, 2013.

[23] D. A. Ketzaki and T. V. Yioultsis, “Metamaterial-Based Design of Planar Compact

MIMO Monopoles,” IEEE Transactions on Antennas and Propagation, vol.61, no.5,

pp. 2758-2766, May 2013.

[24] Y. Lee, D. Ga and J. Choi, “Design of a MIMO Antenna with Improved Isolation

Using MNG Metamaterial,” International Journal of Antennas and Propagation,

vol. 2012, pp.1-7, 2012

234 [25] F. Falcone, T. Lopetegi, J. D. Baena, R. Marques, F. Martin and M. Sorolla,

“Effective negative-/spl epsiv/ stopband microstrip lines based on complementary

split-ring resonators,” in IEEE Microwave and Wireless Components Letters, vol.

14, no. 6, pp. 280-282, June 2004.

[26] R. Selvaraju, M. H. Jamaluddin, M. R. Kamarudin, J. Nasir, and M. H. Dahri,

“Complementary split-ring resonator for isolation enhancement in 5G

communication antenna array,” Progress in Electromagnetics Research, vol. 83, pp.

217–228, 2018.

[27] S. Yoo and S. Kahng, “A Compact MIMO Antenna using ZOR Split Ring Resonator

Radiators with a Decoupling Structure,” Microwave Journal, vol. 54, pp. S26–S31,

2011.

[28] I. O. Vendik and M. Odit, “Theory and Phenomena of Metamaterials,” Metamaterial

Handbook, F. Cappolino (ed.), 2009.

[29] I. J. Bahl and P. Bhartia, “Microwave Solid State Circuit Design,” John Wiley &

Sons, 2003.

[30] C. Caloz, and T. Itoh, “Electromagnetic Metamaterials: Transmission Line Theory

and Microwave Applications,” John Wiley & Sons, 2006.

[31] G. Lubkowski, R. Schumann, and T. Weiland, “Extraction of effective material

parameters by parameter fitting of dispersive models,” Microwave and Optical

Technology Letters, vol. 49, no. 2, pp. 285–288, Jul. 2007.

[32] C. Sabah and S. Uckun, “Multilayer system of Lorentz/Drude type metamaterials

with dielectric slabs and its application to electromagnetic filters,” Progress in

Electromagnetics Research, vol. 91, pp. 349–364, 2009.

235 [33] D. R. Smith, S. Schultz, P. Markos and C. M. Soukoulis, “Determination of Effective

Permittivity and Permeability of Metamaterials from Reflection and Transmission

Coefficients,” Physical Review B, vol. 65, no. 19, 2002, art. 195104.

[34] X. Chen, T. M. Grezegorczyk, B.-I. Wu, J. Pacheco and J. A. Kong, “Robust Method

to Retrieve the Constitutive Effective Parameters of Metamaterials,” Physical

Review E, vol. 70, no. 1, 2004, art. 016608.

[35] J. Zhu, B. Feng, L. Deng and S. Li, “Ultrawideband MIMO/diversity antenna using

CSRR structure for isolation enhancement,” 2015 IEEE 4th Asia-Pacific

Conference on Antennas and Propagation (APCAP), Kuta, 2015, pp. 27-29.

[36] N.A. Jan, et al., (2017) “A Compact CSRR Loaded Monopole Antenna with

Defected Ground Structure for Mobile WLAN and WiMAX Applications,” In:

Otung I., Pillai P., Eleftherakis G., Giambene G. (eds) Wireless and Satellite

Systems. WiSATS 2016. Lecture Notes of the Institute for Computer Sciences,

Social Informatics and Telecommunications Engineering, vol. 186. Springer,

Cham.

[37] D. A. Ketzaki and T. V. Yioultsis, “Metamaterial-Based Design of Planar Compact

MIMO Monopoles,” in IEEE Transactions on Antennas and Propagation, vol.

61, no. 5, pp. 2758-2766, May 2013.

[38] M. S. Sharawi, “Printed MIMO antenna systems: Performance metrics,

implementations and challenges,” Forum for Electromagnetic Research Methods

and Application Technologies (FERMAT), vol. 1, 2014 [Online]. Available:

http://www.e-fermat.org/

236 [39] S. H. Chae, S. Oh and S. Park, “Analysis of Mutual Coupling, Correlations, and

TARC in WiBro MIMO Array Antenna,” in IEEE Antennas and Wireless

Propagation Letters, vol. 6, pp. 122-125, 2007.

[40] C. H. See, R. A. Abd-Alhameed, Z. Z. Abidin, N. J. McEwan and P. S. Excell,

“Wideband Printed MIMO/Diversity Monopole Antenna for WiFi/WiMAX

Applications,” in IEEE Transactions on Antennas and Propagation, vol. 60, no.

4, pp. 2028-2035, April 2012.

[41] Y. Ban, Z. Chen, Z. Chen, K. Kang and J. L. Li, “Decoupled Hepta-Band Antenna

Array for WWAN/LTE Smartphone Applications,” in IEEE Antennas and Wireless

Propagation Letters, vol. 13, pp. 999-1002, 2014.

[42] R. Li, P. Wang, Q. Zheng and R. Wu, “Compact microstrip decoupling and matching

network for two symmetric antennas,” in Electronics Letters, vol. 51, no. 18, pp.

1396-1398, 3 9 2015, doi: 10.1049/el.2015.1738.

[43] C. Xue, X. Y. Zhang, Y. F. Cao, Z. Hou and C. F. Ding, “MIMO Antenna Using

Hybrid Electric and Magnetic Coupling for Isolation Enhancement,” in IEEE

Transactions on Antennas and Propagation, vol. 65, no. 10, pp. 5162-5170, Oct.

2017, doi: 10.1109/TAP.2017.2738033.

[44] P. Liu, D. Sun, P. Wang and P. Gao, “Design of a Dual-Band MIMO Antenna with

High Isolation for WLAN Applications,” Progress In Electromagnetics Research,

74, pp.23-30, 2018.

[45] M. S. Aw, K. Ashwath, and T. Ali. “A compact two element MIMO antenna with

improved isolation for wireless applications,” Journal of Instrumentation 14, no. 06,

2019.

237 CHAPTER SEVEN

Isolation Enhancement of MIMO Antenna Using Defected Ground Structures and Neutralization Line Techniques

The unprecedented evolution of wireless communication systems has raised demands for higher data rates and larger capacity in wireless applications such as Bluetooth, Wi-Fi,

LTE, WLAN and WiMAX. MIMO technology maximizes the benefits of incorporating multiple antennas at the transmitting and receiving ends to achieve higher data rates without consuming extra power or bandwidth if used in a rich scattering environment [1-

3]. However, as they are fitted closely together on confined substrates, the isolation between MIMO antenna elements becomes very low, whereas to achieve the best performance, the isolation must be as high as possible. In this chapter, a compact dual- wideband MIMO antenna is proposed to cover the WLAN bands (2.4/5.2/5.8 GHz) and

WiMAX band (3.5 GHz). The designed antenna basically consists of two uni-planar G- shaped monopole antennas. The technique of integrating two decoupling approaches is applied by combining the effects of the DGS and NL within the proposed MIMO antenna structure.

7.1 The Antenna Layout

The geometry of the proposed MIMO antenna is illustrated in Fig. 7.1. The design is made up of two identical G-shaped slotted radiating elements which are similar to that in

[4]. RT5880LZ substrate that has a thickness of 1.026 mm and dielectric constant (휖r) of

1.96 is used in this design. The total volume of the proposed antenna is 35351.026 mm3, which is suitable for mobile devices. To broaden the antenna bandwidth, two

238 identical defected areas with rectangular shape are made on the ground. The distance between the two radiating elements is set to be 0.144λo at 2.4 GHz (where λo is the free space wavelength). The technique of adding a neutralization line to connect the radiating elements is exploited to reduce the coupling at 2.4 GHz. On the other hand, to minimize the coupling at 3.5 GHz, a rectangular cut is etched in the middle of the ground plane.

Furthermore, stubs are inserted on the feeding lines to further enhance the impedance matching, as shown in Fig. 7.1.

(a) (b)

Fig.7.1. The proposed antenna geometry. (a) Front view; (b) Back view.

7.2 Parametric Study

The primary goal of the parametric studies was to better understand the effects of the defected ground structures, neutralization line and stubs on the MIMO antenna performance. Three different consequences were expected: firstly, to improve the isolation between the radiating elements; secondly, to broaden the antenna bandwidth and finally to enhance the impedance matching. The effects of each decoupling technique will

239 be studied through comparing the S-parameters of the proposed MIMO antenna with and without decoupling technique/techniques.

7.2.1 The Impacts of the DGS Technique on the S-Parameters

The technique of using a defected area in the form of rectangular cut is implemented in this design, with the objectives of increasing the isolation of the antenna and improving the impedance bandwidth [5]. Three defected areas are formed on the bottom layer of the substrate, as shown in Fig. 7.1. The impact of these defected areas on the S-parameters is studied in the following sub-sections.

7.2.1.1 The Influence of the First Defected Area on the S-Parameters

Characteristics

The first case in this parametric study is to examine the effects of the defected areas on the characteristics and performance of the proposed MIMO antenna. The first defected area (DGS1) is cut between the two radiating elements, as shown in Fig. 7.2. The simulated S-parameters of the MIMO antenna with and without DGS1 are illustrated in

Fig. 7.3. It can be clearly seen that the resonant frequency disappears with the existence of the DGS1 within the MIMO antenna structure. On the other hand, the isolation of the proposed antenna is improved around 3.5 GHz with S21 less than 15 dB, as shown in Fig

7.3.

240

Fig.7.2. Schematic structure of the proposed MIMO antenna showing DGS1; (a) Top layer, (b) Bottom layer.

Fig.7.3. Simulated S-parameters of the proposed MIMO antenna with and without DGS1.

7.2.1.2 The Influence of the Second Defected Areas on the S-Parameters

Characteristics

The second case in this parametric study was to evaluate the impact of the twin defected areas under each element (DGS2). The first defected areas are cut underneath the feed

241 line of each radiating element as shown in Fig. 7.4. The simulated S-parameters of the

MIMO antenna with and without DGS2 are shown in Fig. 7.5. It can be clearly seen that the existence of the DGS2 within the MIMO antenna structure helps to enlarge the impedance bandwidth significantly. On the other hand, the isolation of the proposed antenna is improved around 3.5 GHz, with S21 less than 15 dB, as shown in Fig 7.5.

Fig.7.4. Schematic structure of the proposed MIMO antenna showing DGS2; (a) Top layer, (b) Bottom layer.

Fig.7.5. Simulated S-parameters of the proposed MIMO antenna with and without DGS2.

242 7.2.1.3 The Impact of the Three Defected Areas on the S-Parameters

Characteristics

The third case in this parametric study is to evaluate the impact of the defected areas

DGS1 and DGS2 at the same time, as shown in Fig. 7.6. The simulated S-parameters of the MIMO antenna with and without DGS1 and DGS2 are shown in Fig. 7.7. It can be noted that the impedance bandwidth of the proposed MIMO antenna is enhanced with the existence of the three defected areas and it achieves a wide operating band from 2.7 GHz to 5.9 GHz. Moreover, the isolation of the proposed antenna is improved around 3.5 GHz with S21 less than 15 dB and it also shows 0.25 GHz improvement compared to the second case.

Fig.7.6. Schematic structure of the proposed MIMO antenna showing DGS1 and DGS2;

(a) Top layer, (b) Bottom layer.

243

Fig.7.7. Simulated S-parameters of the proposed MIMO antenna with DGS1 and DGS2.

7.2.2 The Influences of the NL on the S-Parameters

The fourth case is to study the effect of inserting NL between the two radiating elements and excluding the effects of DGS1 and DGS2 as shown in Fig. 7.8. The simulated S- parameters of the proposed antenna in the case of the existence and the absence of the neutralization line are demonstrated in Fig. 7.9. It can be clearly seen that the existence of the NL produces a new band around 3.5 GHz and reduces the bandwidth at the first band. On the other hand, the isolation between the antennas with the existence of the neutralization line is improved at the first band and it reaches 15 dB at 2.4GHz.

244

Fig.7.8. Schematic structure of the proposed MIMO antenna with NL and without DGS1 and DGS2; (a) Top layer, (b) Bottom layer.

Fig.7.9. Simulated S-parameters of the proposed MIMO antenna without DGS1 and DGS2 and with NL.

245

The fifth case is to study the effect of combining the DGS decoupling method with the

NL decoupling method in the same proposed antenna structure, as shown in Fig. 7.10.

The simulated S-parameters of the proposed antenna in the case of the existence and the absence of the neutralization line are demonstrated in Fig. 7.11. It can be clearly seen that the existence of the NL produces a new band around 2.4 GHz (from 2.15 to 2.54 GHz) beside the upper band. On the other hand, the isolation between the antennas with the existence of the neutralization line is higher than 15 dB at 2.4 GHz, 3.5 GHz and 5.8 GHz and it is 12 dB at 5.2 GHz.

Furthermore, the effect of the length of the neutralization line is also evaluated. The simulated S-parameters of the antenna with different lengths (L) of the NL are illustrated in Fig. 7.12 and Fig. 7.13 respectively. Fig. 7.12 shows that by increasing the length of the NL, the first band is shifted toward 2.5 GHz while the beginning of the second band is slightly affected. On the other hand, increasing the length of the NL leads to an improvement in the isolation between the antenna elements over the whole band, with considerable enhancement at the first band where the isolation value is greater than, or equal to, 15 dB, as shown in Fig. 7.13.

246

Fig.7.10. Schematic structure of the proposed MIMO antenna showing DGS1, DGS2 and

NL; (a) Top layer, (b) Bottom layer.

Fig.7.11. Simulated S-parameters of the proposed MIMO antenna with DGS1, DGS2 and

NL.

247

Fig.7.12. The simulated reflection coefficient of the proposed MIMO antenna with different lengths for the NL (L1=28 mm, L2=26 mm, L3=24 mm and L4=22 mm).

Fig.7.13. The simulated transmission coefficient of the proposed antenna with different lengths of the NL (L1=28 mm, L2=26 mm, L3=24 mm and L4=22 mm).

248 The surface current distribution of the proposed antenna was used to analyse the contributions of the neutralization line properly with the existence of the DGS areas. Two cases were studied at 2.4 GHz, the first case was the proposed antenna with the neutralization line excluded, while the second case was the proposed antenna including the neutralization line. This study was performed by exciting port 1 and terminating port

2 with a matched load. Fig. 7.14 demonstrates the surface current distribution with and without NL. It can be seen that induced current in the feeding line port 2 is high in the case of the absence of the neutralization line. On the other hand, the existence of the neutralization line leads to the introduction of a new current path which generates an additional coupling to reduce the original coupling, as shown in Fig. 7.14-b [6].

Fig.7.14. Surface current distribution at 2.4 GHz. (a) Without NL, (b) With NL.

249 7.2.3 The Effects of the Stub on the S-Parameters of the

Antenna

To enhance the antenna impedance matching, A stub with an overall length of 5×3.5 mm was added on the feeding line close to the radiating elements [7]. The variations of the S- parameters with and without the stub are shown in Fig. 7.15 and Fig. 7.16 respectively.

It can be clearly seen that the presence of the stub within the antenna structure enhances reflection coefficient at the upper band but decreases it at the end of the upper band at the same time. Two different bands are obtained (2.28–2.58 GHz) and (2.88–5.88 GHz) with

|S11| < -10 dB and the coupling is less than -12 dB. These results indicate that this antenna can cover both WLAN and WiMAX systems.

(dB) with Stub without Stub -10

-20

-30

-40

Reflection Coefficient,S11 Coefficient,S11 Reflection 2 2.5 3 3.5 4 4.5 5 5.5 6 Frequency (GHz)

Fig.7.15. Simulated reflection coefficient S11 with and without neutralization line.

250

0 (dB) with Stub

-5 without Stub S21 S21 -10

-15

Coefficient, -Coefficient, 20

-25 -30 2 2.5 3 3.5 4 4.5 5 5.5 6

Frequency (GHz) Transmission Transmission

Fig.7.16. Simulated transmission coefficient S21 with and without neutralization line.

7.3 Evaluation of Antenna Efficiency and Gain

To assure the best effectiveness of the various decoupling approaches on the proposed

MIMO antenna performance, the antenna radiation efficiency and antenna gain were calculated and compared with the proposed antenna without the decoupling approaches.

The radiation efficiencies and peak gains of the proposed MIMO antenna with and without decoupling techniques over the desired operating bands are shown in Fig. 7.17 and Fig. 7.18 respectively. It is apparent that the antenna with decoupling achieves an efficiency equal to 88% over the first band, an increase of 8% compared to the efficiency of the MIMO antenna without decoupling. On the other hand, the efficiency of the proposed MIMO antenna with decoupling over the second band is also improved and it achieves 87% at 5.5 GHz, an improvement of 4% compared with the efficiency of the

MIMO antenna without decoupling. Further, Fig. 7.18 demonstrates the calculated peak

251 gain of the proposed MIMO antenna over the interval from 2 to 6 GHz. Over the low frequency band, the peak gain of the proposed antenna with decoupling achieves 2 dBi, an improvement of 0.5 dBi compared to the MIMO antenna without decoupling, while the peak gains of the two compared MIMO antennas show good correspondence with each other over the second frequency band with peak gain equal to 4 dBi at 5.5 GHz.

Fig.7.17. Variation of calculated efficiency versus frequency for the proposed MIMO antenna with and without decoupling methods.

Fig.7.18. Simulated peak gain for the proposed MIMO antenna with and without decoupling methods.

252 7.4 Results and Discussions

To validate the simulated findings, a prototype of the proposed MIMO antenna as shown in Fig. 7.19 was manufactured based on the schematic structure in Fig. 7.1 and using the same substrate and dimensions used in the simulation. The measured results were calculated when Element 1 was excited and Element 2 was terminated with a matched load. The scattering parameters were measured using a VNA and the radiation pattern was evaluated in the anechoic chamber at the University of Bradford. In the following subsections, the MIMO antenna performances will be illustrated in detail in terms of S- parameters, ECC, TARC, CCL, DG and 2D and 3D antenna radiation patterns.

(a) (b)

Fig.7.19. The prototype of the proposed antenna, (a) Front view; (b) Back view.

253 7.4.1 S-Parameters of the Antenna

The scattering parameters of the proposed antenna were tested and compared with the corresponding simulated results, as shown in Fig. 7.20 and Fig. 7.21 respectively. It is clearly observed that there is good agreement between the measured and simulated results. It is possible to have slight differences due to the fabrication tolerance, the effects of the connectors and some unspecified losses. As explained in Fig. 7.20, two operating frequency bandwidths appeared, from 2.28 to 2.58 GHz and from 3.0 to 5.88 GHz, with a reflection coefficient S11 < -10 dB and isolation ≥12 dB over the obtained bands.

Therefore, the measured -10 dB frequency bandwidths can well cover the required

WLAN applications.

Fig.7.20. Comparison of reflection coefficient S11 using simulated and measured results.

254

Fig.7.21. Comparison of transmission coefficient S21 using simulated and measured results.

7.4.2 Diversity Performances of the Antenna

The envelope correlation coefficient, channel capacity loss, TARC and diversity gain of the proposed MIMO antenna were investigated in this subsection. These MIMO metrics help to validate the performances within the diversity environment. The ECC was calculated based on the S-parameter equation (3.1). Fig. 7.22 shows the measured results with ρe value below 0.005 which is significantly less than 0.3, the maximum stipulated by the diversity requirements [8]. On the other hand, the simulated and measured channel capacity loss of the proposed design were compared and are shown in Fig. 7.23. In both cases, the CCL does not exceed 1 bps/s/Hz. The simulated and measured average TARCs were compared and illustrated in Fig. 7.24. It is observed that the measured results with

255 TARC value ≤ -10 dB cover the frequency band from 2.25 GHz to 2.7 GHz and from 3.1

GHz to 6 GHz. This indicates that the proposed antenna can be a good candidate to work within WLAN, WiMAX and LTE bands. The differences between the simulated and measured results are attributable to manufacturing and measurement tolerances. Table

7.1 summarizes the characteristics of this proposed MIMO antenna at four frequency bands.

-3 5x 10

Coefficient 2

Correlation Correlation 0 2 2.5 3 3.5 4 4.5 5 5.5 6 Frequency (GHz)

Fig.7.22. The measured envelope correlation coefficient.

Fig.7.23. Simulated and measured channel capacity loss of the proposed antenna.

256

Fig.7.24. Simulated and measured TARC of the proposed antenna.

Table 7.1 The characteristics of the proposed antenna at different frequencies.

MIMO Parameter 2.4 GHz 3.5 GHz 5.2 GHz 5.8 GHz

ECC (10-5) 3.36 0.026 1.32 3.23

Capacity Loss 0.561 0.382 0.543 0.687 (bit/s/Hz)

TARC (dB) -13.84 -18.2 -12.69 -11.9

Finally, the amount of improvement in the diversity antenna system compared to SISO systems can be represented by diversity gain [3]. This metric was calculated depending on equation (3.3). The simulated DG of the proposed MIMO antenna is presented in Fig.

257 7.25 and it can be clearly seen that the DG values for the two frequency bands are very close to the ideal value of DG = 10 dB.

Fig.7.25. The simulated diversity gain of the proposed MIMO antenna.

7.4.3 Radiation Patterns of the Antenna

The far-field patterns of the fabricated design were tested at four selected frequencies,

2.4, 3.5, 5.5, and 5.8 GHz, as depicted in Fig. 7.26. These patterns were tested in the two planes of XZ and YZ, in the case of excitation of port 1 and termination of port 2, and showed a stable omnidirectional pattern at all frequencies. For further evaluation of the volumetric radiation patterns, the three-dimensional variations of the radiated field for the proposed antenna were calculated at four frequency bands and are shown in Fig. 7.27.

This figure gives more appreciation of the field shape as compared to that of the 2-D representations.

258

Fig.7.26. Simulated and measured radiation patterns of the proposed antenna for two planes [(1 and 2) y-x plane and (3 and 4) z-x plane] at (a) 2.4 GHz, (b) 3.5 GHz, (c) 5.5

GHz and (d) 5.8 GHz. Port 1 is excited and port 2 is terminated. Solid curves: simulated results, dashed curves: measured results, “1 and 3” are co-polar components and “2 and

4” are cross-polar components.

259

Fig.7.27. The 3-D patterns of the proposed MIMO antenna at (a) 2.4 GHz, (b) 3.5 GHz,

7.5 Comparison with Published Works

Finally, the proposed MIMO antenna in this chapter is compared with several published data from other researchers. All the MIMO antennas in this comparison used monopole

260 antennas as radiating elements and covered the WLAN band/bands. This comparison is not comprehensive but it gives a fair representative overview of the state of the art of this technology. The comparison is based on the centre frequency of the antenna, the isolation between the radiating elements, the overall electrical size of the antenna, the separation distance between the radiating elements in terms of λo (where λo is the free space wavelength at the lower centre frequency), antenna peak gain, antenna efficiency and

ECC. The summary of this comparison is listed in Table 7.4. It can be seen from the table that the proposed MIMO antenna has the smallest size as compared with the other antennas. This is an important result since antenna miniaturization is highly desirable in many applications. However, the isolation is lower than the others at the second frequency band, and to be more precise, at 5.2 GHz. This comes at the expense of the antenna efficiency where the proposed antenna has the largest efficiency compared with other designs. Moreover, low-level correlation coefficient and good antenna peak gain compared with other designs are also obtained.

Table 7.2. Comparison with other works.

Centre Peak Antenna Isolation Separation Ref. Frequency Electrical Size Gain Efficiency ECC (dB) distance (GHz) (dBi) (%) 2.45 15 1.225λo ×0.817λo 2.1 >55 [9] 0.06λo <0.6 5.48 22 ×0.006λo 1.4 >70 2.5 >10 0.41λo ×0.34 λo 83 [10] 0.081λo <0.006 5.85 >15 ×0.013λo 82 2.48 0.41λo × 0.2λo × [11] >20 0.178λo <0.2 5.06 0.0067 λo 2.39 0.3 λo ×0.28 λo [12] 15 0.0239λo 0.2 5.7 ×0.012 λo 0.68λ ×0.68λ [13] 5.85 >21 o o 4.5 >80 <0.014 ×0.015λo This 2.43 >15 0.27λo×0.27λo×0. 2 88 0.115λo <0.005 work 3.98 >12 008λo 4 87

261 7.6 Conclusions

A compact dual-wideband MIMO antenna for WLAN and WiMAX services has been implemented. The optimised structure consists of two symmetrical G-shaped slotted monopole antennas. A combination of neutralization line and defected ground plane techniques have been used to reduce the coupling and increase the bandwidth. The antenna impedance matching has been improved by printing stubs on each feeding line near to the radiating element. The antenna achieves bandwidths of 12.34% from 2.28 to

2.58 GHz and of 68.49% from 3 to 5.88 GHz, with promising isolation better than 12 dB.

The Envelope correlation coefficient has also been evaluated and found to be less than

0.005 within the obtained frequency bands. Channel capacity loss is another parameter which has been calculated and presented. The obtained value of CCL is less than 1 bps/s/Hz in the two required bands. The simulated and measured average TARCs have also been derived. It has been observed that the measured results with TARC value ≤ -10 dB cover the frequency band from 2.25 GHz to 2.7 GHz and from 3.1 GHz to 6.0 GHz.

Diversity gain is another important metric which has been calculated: the achieved value of DG is very close to the ideal value (10 dB) over the two frequency bands. The simulated and measured radiation patterns were presented at four different frequencies and they showed nearly a stable omnidirectional behaviour. These achievements indicate that the proposed antenna can be a good candidate to work within WLAN, WiMAX and

LTE bands. Some differences have been noticed between the simulated and measured results which can be explained by manufacturing and measurement tolerances.

262 References

[1] Vaughan, R.G. and Andersen, J.B., “Antenna diversity in mobile communications,”

IEEE Transactions on Vehicular Technology, 1987. 36(4): pp. 149-172.

[2] Foschini, G.J. and Gans, M.J., “On limits of wireless communications in a fading

environment when using multiple antennas,” Wireless personal communications,

1998. 6(3): pp. 311-335.

[3] M. R. Sharawi, “Printed MIMO antenna engineering,” Norwood: Artech House, 2014

[4] W. N. W. Marzudi and Z. Z. Abidin, “Dual-wideband G-shaped slotted printed

monopole antenna for WLAN and WiMAX applications," 2013 IEEE International

RF and Microwave Conference (RFM), Penang, 2013, pp. 225-227.

[5] J. Ren, D. Mi and Y.-Z. Yin, “Compact Ultra-wideband MIMO Antenna with

WLAN/UWB Bands Coverage,” Progress in Electromagnetics Research C, vol.50,

pp. 121-129, 2014.

[6] Y. Ban, Z. Chen, Z. Chen, K. Kang and J. L. Li, “Decoupled Hepta-Band Antenna

Array for WWAN/LTE Smartphone Applications,” in IEEE Antennas and Wireless

Propagation Letters, vol. 13, pp. 999-1002, 2014.

[7] R. A. Bhatti, J. Choi and S. Park, “Quad-Band MIMO Antenna Array for Portable

Wireless Communications Terminals,” in IEEE Antennas and Wireless Propagation

Letters, vol. 8, pp. 129-132, 2009.

[8] J. Li, Q. Chu and T. Huang, “A Compact Wideband MIMO Antenna with Two Novel

Bent Slits,” in IEEE Transactions on Antennas and Propagation, vol. 60, no. 2, pp.

482-489, Feb. 2012.

263 [9] R. Addaci, A. Diallo, C. Luxey, P. Le Thuc and R. Staraj, “Dual-Band WLAN

Diversity Antenna System with High Port-to-Port Isolation,” in IEEE Antennas and

Wireless Propagation Letters, vol. 11, pp. 244-247, 2012.

[10] Y. Ou, X. Cai and K. Qian, “Two-element compact antennas decoupled with a simple

neutralization line,” Progress in Electromagnetics Research, vol. 65, pp. 63-68,

2017.

[11] C. Wang, H. Zhou and Y. Song, “Design of Low Coupling MIMO Antenna Array

with Parasitic Ground Structure,” 2018 International Conference on Microwave and

Millimeter Wave Technology (ICMMT), Chengdu, 2018, pp. 1-3, doi:

10.1109/ICMMT.2018.8563774.

[12] A. Dkiouak, A. Zakriti, M. E. Ouahabi and A. Mchbal, “Design of a two symetrical

F-shaped MIMO antenna for Wi-MAX and WLAN applications,” 2019

International Conference on Wireless Technologies, Embedded and Intelligent

Systems (WITS), Fez, Morocco, 2019, pp. 1-6, doi: 10.1109/WITS.2019.8723859.

[13] A. Kumar Saurabh, P. Singh Rathore and M. Kumar Meshram, “Compact wideband

four-element MIMO antenna with high isolation,” in Electronics Letters, vol. 56, no.

3, pp. 117-119, 6 2 2020, doi: 10.1049/el.2019.2871.

264 CHAPTER EIGHT

Conclusions and Future Work

Recently, increases in the data rate and the reliability of wireless communication systems have become high priorities. Therefore, MIMO technology has been applied to achieve better performance by utilizing multiple channels between the transmitter and receiver in the multi-path environment. Multiple radiating elements must be used to establish multiple channels in the transmitter and receiver system ends. These elements should be independent of each other in order to increase the isolation between them and their associated channels. This can be achieved by increasing the distance between the radiating elements, but it is not a practical solution to put them far away from each other due to the high demand for compact devices in communication systems. Consequently, many techniques have been developed in the last ten years to improve the isolation between the radiating elements in narrow spaces.

The potentials of integrating different mutual coupling reduction methods have been investigated in this thesis through the design and implementation of several MIMO antennas with different geometries to work at multi-band frequencies and cover the main applications of wireless communication systems. The usefulness of these methods has been illustrated, based on the characteristic performance analysis of the proposed MIMO antenna systems. Hence the conclusions of this project may be summarized as follows:

Chapter Two: the literature review of the MIMO antenna system has been introduced in this chapter. The history of MIMO antennas was reviewed with a brief comment about

265 the developments which led to an improvement in its diversity and an increase in the average data rates. Then, mutual coupling between MIMO antenna elements has also been discussed as the primary reason for the deterioration in the system performance.

After that, several popular mutual coupling reduction techniques, such as antenna placement and orientation, metamaterial resonators, parasitic elements, neutralization lines, defected ground structures and hybrid decoupling method have been addressed, including the implementation of these techniques on several antenna structures. The developed structures help to illustrate the recent directions of this research field. Different performance metrics related to MIMO antenna systems have been discussed to show their effect on MIMO antenna performance. The level of the isolation between the elements shows how much the elements are isolated within the configuration of the antenna while the radiation patterns of the antennas depend on the value of the correlation coefficients to specify how much they are isolated in the environment. On the other hand, the TARC value is used to calculate the amount of radiated power to the total available power.

Channel capacity shows the improvement value in the data rate. Finally, the diversity gain is a significant factor which indicates how much the diversity affects the performance of the communication system.

Chapter Three: A compact dual-band MIMO antenna operating at WLAN bands

(2.4/5.2/5.8 GHz) has been designed. The schematic structure of this antenna consists of two double T-shaped radiating elements with an integrated decoupling method that combines DGS and neutralization line. The optimum value of isolation has been obtained by optimizing the dimensions of the decoupling structures. The results demonstrate that

266 DGS has increased the bandwidth and reduced the coupling at the second band while applying NL has helped to increase the isolation to more than 15 dB at the first band.

Chapter Four: A low–profile tri-band MIMO antenna operating in WLAN (2.4/5.2/5.8

GHz) and WiMAX (3.5 GHz) bands has been designed. This antenna employs two printed uni-planar monopole antennas. A new hybrid technique has been applied to improve the isolation between the radiating elements by introducing DGS and neutralization on ground planes. The main benefit of this technique is to reduce the coupling caused by the surface current in the ground plane. The outcomes indicate that applying the hybrid decoupling approach has helped to improve the isolation to more than

15 dB at the three required bands.

Chapter Five: A compact dual-band MIMO antenna operating at WLAN bands

(2.4/5.2/5.8 GHz) has been designed. This antenna has two concentric double square rings as radiating elements, printed symmetrically on the same substrate. A new method that combines DGS with five parasitic elements has been introduced to reduce the unwanted coupling between the radiating elements. Results show that applying the DGS decoupling approach has improved the antenna isolation to more than 15 dB at the second band whilst the parasitic method enhanced the isolation at the first band.

Chapter Six: The antenna placement and orientation decoupling approach, alongside metamaterial-based isolation enhancement structures have been investigated in this chapter. A 2×2 MIMO antenna consisting of two dual-band arc-shaped radiating elements working in WLAN and sub-6 GHz 5th generation (5G) bands has been modelled and implemented. Four complementary split-ring resonators (SRRs) generating negative

267 permeability and permittivity have been added between the radiating elements to reduce the coupling between the elements. In this design, applying the first decoupling method has helped to improve the isolation in the second band of operation whilst the second method has enhanced the isolation in the first band. A good diversity improvement with low ECC has been obtained.

Chapter Seven: A compact dual-wideband MIMO antenna for WLAN and WiMAX services has been implemented. The optimised structure consists of two symmetrical G- shaped slotted patch antennas. A combination of neutralization line and defected ground plane techniques has been used to reduce the coupling and increase the bandwidth. The antenna impedance matching has been improved by printing stubs on the feeding lines near to the radiating elements. Analysing the obtained results shows that inserting NL and DGS have helped to increase the isolation to more than 15 dB around 2.45 GHz and

3.5 GHz.

8.1 Future Work

The plan for future works can be used to extend the directions of the work achieved. The following examples are the most interesting themes that require further investigation in the future:

• The design of Ultra-Wideband (UWB) antenna arrays is a very interesting area

for researchers to obtain more significant design strategies to fit in compact and

limited 3D areas. The obtained findings of this work could be applied in UWB

radar applications for breast cancer tumour detection [1].

268

• Another subject related to human safety is the reduction of specific absorption

rate (SAR) in the tissues of the human head or hands. The concept of the required

design is by using a MIMO antenna in the millimetre band to minimize the

radiated power in the near field [2].

• Another area of interest is using a Genetic Algorithm (GA) to evaluate the effect

of neutralization lines within the MIMO antenna structure and to derive

unconventional and non-intuitive designs. This can be applied to complex antenna

systems.

• Investigation of the massive MIMO antenna technology is a fruitful avenue for

future research: it is highly appropriate the 5th and 6th generation mobile

applications due to its high data rate and spectral efficiency [3].

• Beam-forming in the millimetre waves is also an attractive area of research. This

technique uses the advantages of MIMO wireless systems and can be considered

another key to future expansion of 5G and 6G technologies [4].

269 References

[1] H. Song et al., “Detectability of Breast Tumors in Excised Breast Tissues of Total

Mastectomy by IR-UWB-Radar-Based Breast Cancer Detector,” in IEEE

Transactions on Biomedical Engineering, vol. 66, no. 8, pp. 2296-2305, Aug. 2019.

[2] M. E. Yanik and M. Torlak, “Near-Field MIMO-SAR Millimeter-Wave Imaging

with Sparsely Sampled Aperture Data,” in IEEE Access, vol. 7, pp. 31801-31819,

2019.

[3] Y. Li, C. Sim, Y. Luo and G. Yang, “High-Isolation 3.5 GHz Eight-Antenna MIMO

Array Using Balanced Open-Slot Antenna Element for 5G Smartphones,” in IEEE

Transactions on Antennas and Propagation, vol. 67, no. 6, pp. 3820-3830, June

2019.

[4] B. Bahreini, H. Oraizi, N. Noori and P. Mousavi, “Optimum Design of a Beam-

Forming Array of S-Shaped DRA Elements with a Superstrate on an SIW Feed for

5G Mobile Systems,” in IEEE Antennas and Wireless Propagation Letters, vol. 18,

no. 7, pp. 1410-1414, July 2019.

270