An Evaluation of Orbital Angular Momentum Multiplexing Technology

applied sciences

Article An Evaluation of Orbital Angular Momentum Multiplexing Technology

Doohwan Lee *, Hirofumi Sasaki, Hiroyuki Fukumoto, Yasunori Yagi and Takashi Shimizu

NTT Network Innovation Laboratories, NTT Corporation, 1-1 Hikarinooka, Yokosuka-Shi 239-0847, Japan; [email protected] (H.S.); [email protected] (H.F.); [email protected] (Y.Y.); [email protected] (T.S.) * Correspondence: [email protected]; Tel.: +81-46-859-4778

Received: 8 March 2019; Accepted: 24 April 2019; Published: 26 April 2019

Abstract: This paper reports our investigation of wireless communication performance obtained using orbital angular momentum (OAM) multiplexing, from theoretical evaluation to experimental study. First, we show how we performed a basic theoretical study on wireless OAM multiplexing performance regarding modulation, demodulation, multiplexing, and demultiplexing. This provided a clear picture of the effects of mode attenuation and gave us insight into the potential and limitations of OAM wireless communications. Then, we expanded our study to experimental evaluation of a dielectric lens and end-to-end wireless transmission on 28 gigahertz frequency bands. To overcome the beam divergence of OAM multiplexing, we propose a combination of multi-input multi-output (MIMO) and OAM technology, named OAM-MIMO multiplexing. We achieved 45 Gbps (gigabits per second) throughput using OAM multiplexing with five OAM modes. We also experimentally demonstrated the effectiveness of the proposed OAM-MIMO multiplexing using a total of 11 OAM modes. Experimental OAM-MIMO multiplexing results reached a new milestone for point-to-point transmission rates when 100 Gbps was achieved at a 10-m transmission distance.

Keywords: orbital angular momentum multiplexing; OAM; OAM-MIMO; 28 GHz; uniform circular array; dielectric lens

1. Introduction Recently, wireless communication using OAM (Orbital Angular Momentum) has drawn much attention as an emerging candidate for beyond 5G (fifth generation) technology due to its potential as a means to enable high-speed wireless transmission. OAM is a physical property of electro-magnetic waves that are characterized by a helical phase front in the propagation direction. Since the characteristic can be used to create multiple independent channels, wireless OAM multiplexing can effectively increase the transmission rate in a point-to-point link such as wireless backhaul and/or fronthaul [1,2]. Recent seminar work demonstrated the feasibility of OAM multiplexing by achieving 32 Gbps (gigabits per second) transmission, as Yan et al. reported using the 28 GHz (gigahertz) band in 2014 [1] and the 60 GHz band in 2016 [3]. Since OAM multiplexing technology is relatively new, it is important to validate the feasibility from various perspectives. To do that, we first validated the feasibility from a theoretical perspective using simulations (Section2). We then validated the feasibility from beam generation and propagation perspectives in experiments (Section3). Finally, we concluded by validating the feasibility from the end-to-end wireless communication perspective using experiments (Section4). In our previous research, we explored the potential of wireless OAM multiplexing by conducting the following three studies. The first part of our work was theoretically investigating the feasibility of OAM multiplexing. First, we investigated the theoretical performance of modulation, demodulation, multiplexing,

Appl. Sci. 2019, 9, 1729; doi:10.3390/app9091729 www.mdpi.com/journal/applsci Appl. Sci. 2019, 9, 1729 2 of 13 and demultiplexing OAM algorithms using computer simulations [4]. This enabled us to clarify the performance and effect of mode-dependent power attenuation. In doing so we generated OAM signals by using a UCA (uniform circular array) that comprises multiple omnidirectional antenna elements. The second part of our work was validating the feasibility from beam generation and propagation perspectives. In particular, we expanded our study to the usage of a dielectric lens to examine the feasibility of long distance transmission using OAM [5]. Ideally, all OAM modes are orthogonal to each other due to the unique nature of their phase fronts. However, with these modes, it is difficult to transmit over long distances because their energy rapidly diverges as the beam propagates. To achieve long-distance transmission, we developed and proposed a beam divergence reduction method using the focusing effect of a dielectric lens. We conducted a wave propagation experiment on 28 GHz bands to demonstrate the effectiveness of using such a lens. In the experiment, we were able to generate OAM modes 0, 1, and 2. In addition, we showed the beam divergence reduction effect by making a ± ± comparison between conventional OAM beam generation methods and using a dielectric lens with a UCA. In the third part of our work, we validated the feasibility from the end-to-end wireless communication perspective. We report experimental results using wireless OAM multiplexing at 28 GHz. One of the major challenges for achieving OAM multiplexing is its intensity variation among different OAM modes. As shown in Figure 7, intensity distributions of OAM signals with different modes are given by the Bessel function of the first kind by their inherent nature [2]. This yields selective reception (Rx) SNR (signal-to-noise ratio) degradation when the Rx antenna is located in a low SNR region (null region). To address this problem, we used multiple UCAs, which are designed to avoid reception in null regions. We obtained successful experimental results using five OAM modes over 28 GHz [6]. We also developed and proposed OAM-MIMO multiplexing using multiple UCAs. Unlike OAM multiplexing, OAM-MIMO multiplexing exploits multiple sets of the same OAM modes with receiver equalizations [7]. This enables the number of concurrently transmitted data streams to be increased without using higher OAM modes that have large beam divergence. We experimentally demonstrated the effectiveness of the proposed OAM-MIMO multiplexing by using 11 OAM modes in total (three OAM modes 0 and two sets each of OAM modes 1 and 2). Experimental results ± ± reached a new milestone in point-to-point transmission rates when 100 Gbps was achieved at 10 m transmission distance.

2. Background and Theoretical Performance Evaluation

2.1. OAM Generation Using a Uniform Circular Array Studies regarding OAM multiplexing in the wireless communication ﬁeld are categorized into antenna design and beam generation, end-to-end experiments, signal processing methods, and system studies for topics such as capacity analysis and link budget. Among these, we mainly focus on antenna design in this subsection. Various types of antenna designs have been reported using helicoidally deformed parabolic antennas [8], spiral phase plates (SPP) [2,3,9], holographic plates [10], elaborately tuned planer SPPs [11], and other components [12,13]. In our latest research, we focused on OAM generation using UCAs [14–16]. Figure1 shows OAM mode generation by using UCAs. The phase of each antenna is shifted in accordance with an OAM mode. The transmitted signal from each antenna can be written in vector form as

L 2π(N 1) T j 2π j − x = 1, e N , , e N , (1) ··· where L is the OAM mode number and N is the number of radiating antennas in the transmitting UCA. By using a UCA, the OAM state number L is limited by the number of transmitting antennas as |L| < N/2. Appl. Sci. 2019, 9, 1729 3 of 13 Appl. Sci. 2019, 9, x FOR PEER REVIEW 3 of 13

FigureFigure 1. GenerationGeneration of of OAM OAM modes modes by by a a uniform uniform circular circular array. Since the diffraction pattern of the UCA can be approximated by the Bessel beam [17], the field Since the diffraction pattern of the UCA can be approximated by the Bessel beam [17], the field distribution of the beam with the OAM mode L is often expressed in the Bessel beam equation as distribution of the beam with the OAM mode L is often expressed in the Bessel beam equation as ! λ exp[(2πi/λ) √2r2 +2z2] λ exp[(2πi / λ) r + z ] − L  2π2rDπrD  vL(r,()θ,θz) == ⋅ iL− exp[][θiLθ⋅ ] J  , (2) vL r, , z i exp iL J L L , (2) 4π √r22 + z22 · ·  λ √2 r2 +2 z2 4π r + z  λ r + z  where JL( ),⋅λ, and D, respectively, denote the Lth order Bessel function of the first kind, the wavelength where J L· ( ) , λ, and D, respectively, denote the Lth order Bessel function of the first kind, the of the carrier frequency, and the radius of the transmitting UCA. Equation (2) is represented in wavelength of the carrier frequency, and the radius of the transmitting UCA. Equation (2) is cylindrical coordinates, where r and θ are, respectively, radius and azimuthal angle at the Rx plane represented in cylindrical coordinates, where r andθ are, respectively, radius and azimuthal angle at that is vertical to the beam propagation direction. z is the distance between the centers of the Tx the Rx plane that is vertical to the beam propagation direction. z is the distance between the centers (transmission) and Rx UCAs. Here, let us discuss a comparison between OAM and MIMO. OAM is of the Tx (transmission) and Rx UCAs. Here, let us discuss a comparison between OAM and MIMO. a specific implementation of MIMO with circular antenna arrays. The difference is that an OAM OAM is a specific implementation of MIMO with circular antenna arrays. The difference is that an beam can be achieved virtually and practically in accordance with an OAM state, so it is possible to OAM beam can be achieved virtually and practically in accordance with an OAM state, so it is design a mechanism to manipulate that beam specifically to achieve a specific type of communication. possible to design a mechanism to manipulate that beam specifically to achieve a specific type of The difference from the conventional MIMO and OAM multiplexing is as follow. To obtain a full rank communication. The difference from the conventional MIMO and OAM multiplexing is as follow. To matrix, usually an NLOS (non-line-of-sight) multipath and a rich scattering environment are assumed obtain a full rank matrix, usually an NLOS (non-line-of-sight) multipath and a rich scattering in the conventional MIMO. In LOS (line-of-sight) environment cases, digital signal processing such as environment are assumed in the conventional MIMO. In LOS (line-of-sight) environment cases, precoding at the transmitter may be required for the conventional MIMO case to obtain a full rank digital signal processing such as precoding at the transmitter may be required for the conventional matrix. On the other hand, OAM does not need digital signal processing at the transmitter to obtain MIMO case to obtain a full rank matrix. On the other hand, OAM does not need digital signal the full rank matrix. processing at the transmitter to obtain the full rank matrix. 2.2. Modulation and Demodulation 2.2. Modulation and Demodulation Modulation and demodulation using OAM can be mainly categorized into two schemes. We detail theseModulation schemes as follows.and demodulation using OAM can be mainly categorized into two schemes. We detail these schemes as follows. • OAM Shift Shift Keying Keying (OAMSK) (OAMSK) [14]: [14]: This This scheme scheme simply simply puts puts binary binary data data into into an OAM an OAM mode. mode. For • example,For example, bit “0” bit “0”is mapped is mapped as OAM as OAM mode mode 1, 1,while while bit bit “1” “1” is ismapped mapped as as mode mode -1 -1 (minus (minus 1). 1). OAMSK modulated signals can be demodulated by using the phase gradient method, an FFT (fast FourierFourier transform)transform) basedbased method,method, oror MLML (maximum(maximum likelihood)likelihood) detection.detection. The gradient method uses the phase difference difference between two receiving receiving antennas antennas to to determine determine the the OAM OAM mode. mode. The FFT-based method conducts the FFT proces processs using a reception (Rx) UCA and chooses the maximum coefficients. coefficients. ML ML detection detection selects selects the the OAM OAM mode mode with with the the closest distance to the received signal. • OAM Division Multiplexing (OAMDM) [16]: [16]: This This scheme uses OAM modes to carry multiple • streams of data simultaneously. An An OAM mode mode ca cann carry one stream, similar to the way that one OFDM (orthogonal frequency division multiplexing) subcarrier can. This This scheme scheme potentially potentially improves the spectrum efficiency. efficiency. With it, OA OAMDMMDM modulated signals are demodulated similar to the the way way they they are are with with MIMO MIMO equalization equalization techniques techniques such such as zero as zeroforcing forcing or minimum or minimum mean squaremean square error errorequalization, equalization, assuming assuming the thechannel channel information information is isavailable. available. Since Since OAM multiplexing isis expected expected to to be be used used under under LOS LOS environments environments with staticwith channelsstatic channels such as wirelesssuch as wirelessfronthaul fronthaul/backhaul,/backhaul, simplified simplified channel estimationchannel estimation using Equation using Equation (2) might (2) be might feasible. be feasible.

2.3. Mode-Dependent Power Distribution Appl. Sci. 2019, 9, 1729 4 of 13

2.3. Mode-Dependent Power Distribution In the work we report here, we also considered two key issues regarding the mode-dependent powerAppl. distribution Sci. 2019, 9, x FOR among PEER REVIEW diﬀerent OAM modes. These issues are as follows. 4 of 13

PeakIn Rx the Power work we Degradation: report here, we As also the considered number of two OAM key modesissues regarding increases, the the mode radiation-dependent becomes • powerwider, distribution the angle from among the different beam axis OAM at themodes. peak These Rx power issues ar becomese as follows. wider, and the SNR at its peak Rx powerPeak Rx becomes Power Degradation smaller. Accordingly,: As the number the of performance OAM modes is increases, degraded the as radiation the number becomes of OAM modeswider, increases. the angle from the beam axis at the peak Rx power becomes wider, and the SNR at its peak Rx power becomes smaller. Accordingly, the performance is degraded as the number of Non-identical Peak Rx Power Locations: The peak Rx power locations of each OAM mode • OAM modes increases. are notNon identical-identical Peak because Rx Power their radiationLocations: patternsThe peak areRx power distinct. locations Therefore, of each the OAM mode-dependent mode are performancenot identical degradation because their becomes radiation more patterns severe are when distinct. a single Therefore, Rx UCA the is mode used-dependent because some OAMperformance modes might degradation not have becomes the peak more Rx powersevere when at a certain a single location. Rx UCA is used because some OAM modes might not have the peak Rx power at a certain location. 2.4. Evaluation 2.4 Evaluation We implemented a simulation testbed of OAM based wireless communication at 60 GHz. Figure1 shows anWe illustration implemented of the a simulation generation testbed of OAM of OAM signals base usingd wireless UCA. communication The gain for at each 60 GHz. antenna Figure element reﬂects1 shows the UCAan illustration as set to of be the 0 generationdBi (decibels of OAM relative signals to isotropicusing UCA. radiator). The gain Note for each that antenna concurrent transmissionelement reflects of multiple the UCA OAM as modesset to be can 0 be dBi achieved (decibels by relative superposing to isotropic multiple radiator) OAM. Note signals. that It is concurrent transmission of multiple OAM modes can be achieved by superposing multiple OAM generally assumed that OAM multiplexing is to be mainly used in LOS environments such as signals. It is generally assumed that OAM multiplexing is to be mainly used in LOS environments wireless fronthaul and/or backhaul. Therefore, we used an AWGN (additive white Gaussian noise) such as wireless fronthaul and/or backhaul. Therefore, we used an AWGN (additive white Gaussian channelnoise) environment. channel environment.

(a) (b)

FigureFigure 2. Performance 2. Performance evaluations evaluations of:of: (a) OAMSK (phased (phased gradient gradient method method with with varying varying Rx array Rx array radius);radius) (b); OAMSK (b) OAMSK (FFT-based (FFT-based and and ML methods);ML methods) (c); OAMDM (c) OAMDM with with ﬁxed fixed Rx power; Rx power and; and(d) OAMDM (d) withOAMDM varying Rxwith array varying radius. Rx array radius. Appl. Sci.Appl. 2019 Sci., 9,2019 x FOR, 9, PEER 1729 REVIEW 5 of 13 5 of 13

Figure 2a shows the OAMSK performance obtained by the phase gradient method, which uses two antennasFigure for2 OAMa shows signal the OAMSKdetection performancewith varying obtained Rx array byradius thephase and angular gradient distance method, between which uses two Rxtwo antenna antennas elements. for OAM Although signal detection the performanc with varyinge is poorer Rx array than radius that obtained and angular with distance FFT-based between and MLtwo methods, Rx antenna only elements. two antenna Although elements the performance are necessary. is poorerThis is than in contrast that obtained to cases with in which FFT-based all and the RxML UCA methods, elements only are two required. antenna This elements is favora areble necessary. for higher This OAM is in mode contrast transmission. to cases in which Figure all 2b the Rx showsUCA the OAMSK elements performance are required. obtained This is favorable by using forthe higher FFT-based OAM method mode transmission. and ML detection Figure while2b shows varyingthe the OAMSK number performance of antenna obtained components by using in the the Rx FFT-based UCA. We method found andthat MLin both detection cases while the ML varying detectionthe numberyields generally of antenna better components results and in additi the Rxonal UCA. performance We found gain that is in achieved both cases as thethe MLnumber detection of antennayields components generally better in the results Rx UCA and additionalincreases. performanceFigure 2c shows gain the is achievedOAMDM as performance the number ofwith antenna fixed componentstotal Rx power in the among Rx UCA OAM increases. modes. FigureNote 2thatc shows Rx signals the OAMDM in this performancecurve are obtained with fixed at the total Rx locationpower of the among peak OAMRx power modes. of each Note mode. that Rx In signalsthis case, in thiswe observed curve are performance obtained at the degradation location of of the 3 peak dB as Rxthe power number of of each OAM mode. modes In this increased. case, we Figure observed 2d shows performance the effect degradation of a non-identical of 3 dBas location the number of of the peakOAM Rx modespower increased.by varying Figure the Rx2 darray shows radius. the e Asffect the of OAM a non-identical mode number location increases, of the peakthe Rx Rx array power by radiusvarying for the thebest Rx BER array (bit radius. error Asrate) the performanc OAM modee also number increases. increases, Correspondingly, the Rx array radius if the for Rx the array best BER radius(bit is customized error rate) performance for a certain also OAM increases. mode, th Correspondingly,e performance of if other the Rx OAM array modes’ radius signals is customized might for a be deterioratedcertain OAM severely. mode, For the this, performance multiple UCAs of other might OAM be modes’necessary, signals which might leads be us deteriorated to use multiple severely. UCAs.For this, multiple UCAs might be necessary, which leads us to use multiple UCAs. In thisIn subsection, this subsection, we wereport report how how we we studied studied the the potential andand limitations limitations of of OAM-based OAM-based wireless wirelesscommunication communication in termsin terms of mode-dependentof mode-dependent power power attenuation attenuation and and non-identical non-identical peak peak Rx Rx power powerlocations locations through through the the use use of of modulation modulation and and demodulation demodulation algorithms. algorithms. We We confirmed confirmed the the effect of effect mode-dependentof mode-dependent performance performance variations. variations. Further Further study study is necessary is necessary to fully to rectifyfully rectify the undesirable the undesirableeffects ofeffects the mode-dependent of the mode-dependent power attenuation.power attenuation.

3. Beam3. BeamFocusing Focusing Effect E Usingﬀect Using Dielectric Dielectric Lens Lensfor OAM for OAM Multiplexing Multiplexing

3.1. Usage of Dielectric Lens for OAM Multiplexing 3.1. Usage of Dielectric Lens for OAM Multiplexing We presentWe present the beam the beam focusing focusing effect eobtainedﬀect obtained by using by using a dielectric a dielectric lens for lens OAM for OAM multiplexing multiplexing to increaseto increase the transmission the transmission distance. distance. Figure Figure 3 shows3 shows the configuration the conﬁguration of the of transmitting the transmitting and and receivingreceiving antennas. antennas. The OAM The OAM mode mode radiates radiates from from the theUCA UCA installed installed on on the the back back of of the the dielectric dielectric lens. lens. TheThe OAMOAM mode mode radiated radiated into into the spacethe space is phase-modulated is phase-modulated by the lensby the and lens reaches and the reaches reception the point. n receptionWhen point. transmitting When transmitting the OAM mode the OAM, we mode weight n, we each weight element each of element the array of antenna. the array Since antenna. the phase Since distributionthe phase distribution in the OAM in mode the OAM is circularly mode is symmetric circularly in symmetric the plane perpendicularin the plane perpendicular to the beam traveling to the beamdirection, traveling a circular direction, lens isa usedcircular sothat lens circular is used symmetry so that circular is notdisturbed symmetry by is phase not disturbed modulation by at the phasetime modulation of passing at the through time theof passing lens. through the lens.

Figure 3. Configuration of the Tx and Rx antennas using dielectric lens. Figure 3. Configuration of the Tx and Rx antennas using dielectric lens. By properly setting focal length f of the lens considering the distance between the lens and Bythe properly transmitting setting antenna, focal length the beam f of the can lens be considering narrowed by the the distance light converging between the eff ect.lens Forand example,the transmittingFukumoto antenna, et al. [4 ,the18] showedbeam can that be the narrowed beam spread by canthe be light reduced converging based on effect. the imaging For example, magnification Fukumotoat the receptionet al. [4,18] point. showed that the beam spread can be reduced based on the imaging magnification at the reception point. Appl. Sci. 2019, 9, 1729 6 of 13

Appl. Sci. 2019, 9, x FOR PEER REVIEW 6 of 13 3.2. Experiments of OAM Multiplexing Using Dielectric Lens 3.2. Experiments of OAM Multiplexing Using Dielectric Lens To confirm the basic operation produced in using a dielectric lens, we performed a propagation experimentTo confirm using the the basic 28 GHzoperation band produced in an anechoic in using chamber. a dielectric Figure lens,4 shows we performed the experimental a propagation setup. experimentThe measurement using the frequency 28 GHz of band the network in an anechoic analyzer chamber. (NWA) wasFigure set from4 shows 2 to the 3 GHz. experimental Radio frequency setup. The(RF) measurement chains up-converted frequency 28–29 of GHz the insertednetwork signalsanalyzer and (NWA) fed them was to theset Txfrom antenna. 2 to 3 At GHz. the receiver Radio frequencyside, the received (RF) chains signals up-converted were down-converted 28–29 GHz toinsert 2–3ed GHz signals signals and by fed the them RF chains to the attached Tx antenna. directly At theafter receiver the Rx antenna.side, the Wereceived used asignals millimeter were wavedown commercial-converted to lens 2–3 whose GHz directionalsignals by gainthe RF and chains beam attachedhalf-width directly were, after respectively, the Rx antenna. 35 dBi andWe used 3 cm a (centimeters). millimeter wave The commercial design requirements lens whose for directional dielectric gainlenses, and including beam half-width focal length, were, refractive respectively, index, 35 diameter, dBi and directional3 cm (centimeters). gain, and The beam design half-width, requirements remain foropen dielectric for analysis. lenses, including focal length, refractive index, diameter, directional gain, and beam half-width,In the propagationremain open measurements,for analysis. we measured the channel coefficient between the Tx and Rx antennasIn the using propagation NWA while measurements, operating the we positioner measured to movethe channel the Tx coefficient and Rx antennas between to form the Tx the and Tx andRx antennasRx UCAs using elements’ NWA location while operating sequentially. the In positioner other words, to move we emulated the Tx and Tx Rx and antennas Rx UCAs to using forma the single Tx and Rx UCAs elements’ location sequentially. In other words, we emulated Tx and Rx UCAs using a Tx and Rx antenna by sequentially measuring the channel coefficients of two points and combined single Tx and Rx antenna by sequentially measuring the channel coefficients of two points and the entire measurements to obtain channel characteristics between the emulated Tx and Rx antennas. combined the entire measurements to obtain channel characteristics between the emulated Tx and The gain for both Tx and Rx antennas element was 27 dBi. We obtained M N channel matrix H with Rx antennas. The gain for both Tx and Rx antennas element was 27 dBi. We× obtained M × N channel measured channel coefficients consisting of all combinations between N-points (Tx side) and M-points matrix H with measured channel coefficients consisting of all combinations between N-points (Tx (Rx side). Subsequently, we performed a matrix operation corresponding to OAM mode generation side) and M-points (Rx side). Subsequently, we performed a matrix operation corresponding to OAM on the measured channel matrix and evaluated the phase and intensity distributions. The amplitude mode generation on the measured channel matrix and evaluated the phase and intensity and phase characteristics of the OAM mode n were obtained by multiplying the column vector that distributions. The amplitude and phase characteristics of the OAM mode n were obtained by generates OAM mode n by obtained channel matrix H. multiplying the column vector that generates OAM mode n by obtained channel matrix H.

FigureFigure 4. 4. ExperimentalExperimental environment environment (OAM (OAM mult multiplexingiplexing using using dielectric dielectric lens). lens). Table1 shows experimental parameters. First, to conﬁrm that an OAM mode was generated at Table 1. Experimental parameters (Dielectric lens). the Rx side, we measured the phase distribution and the intensity distribution formed by the beam passing through the lens on the Rx side.Parameter Value Focal length: f 0.30 m LensTable 1. ExperimentalDiameter: parameters D (Dielectric lens). L 0.30 m NumberParameter of antenna elements Value 12 UCA Diameter:Focal length: Df 0.300.04 mm Lens T Distance betweenDiameter: lensD Land UCA: a 0.300.40 m m NumberFrequency of antenna elements28 12GHz Others UCA Distance betweenDiameter: TxDT and Rx: b 0.043.12 m m Distance between lens and UCA: a 0.40 m

Frequency 28 GHz Table 1 shows experimentalOthers parameters. First, to confirm that an OAM mode was generated at Distance between Tx and Rx: b 3.12 m the Rx side, we measured the phase distribution and the intensity distribution formed by the beam passing through the lens on the Rx side.

Appl. Sci. 2019, 9, 1729 7 of 13

Figure5 shows the measurement results and simulation results of the intensity distribution and phase distribution of OAM modes 2, 1, 0, +1, and +2. The measurements were conducted over a − − 60 cm 60 cm grid and data were acquired at 3 cm intervals. The ﬁgure results conﬁrmed that OAM × beams whose phases rotate as much as their mode order were obtained in the experiments. The results also well matched the simulation results. Further, we consider that our experimental method will makeAppl. it possible Sci. 2019,, 9 to,, xx FORFOR emulate PEERPEER REVIEWREVIEW OAM beam generation. 7 of 13

FigureFigure 5. Measurement 5. Measurement results results and and simulation simulation results of of phase phase and and intensity intensity distributions. distributions.

Next,Figure we conducted 5 shows the experiments measurement to results ascertain and thesimula beamtion divergence results of the reduction intensity distribution eﬀect obtained and with the lens.phase We distribution took into accountof OAM themodes distance -2, -1, from0, +1, theand center+2. The of measurements the Rx side to were the locationconducted of over the strongesta 60 beamcm intensity × 60 cm since grid itand is thedata quantitative were acquired index at 3for cm evaluating intervals. The beam figure divergence. results confirmed We deﬁned that thisOAM metric as thebeams beam whose diameter. phases The rotate smaller as much is the as beam their diameter, mode order the were smaller obtained is the in beam the experiments. divergence atThe the Rx results also well matched the simulation results. Further, we consider that our experimental method side because the beam energy is more focused close to the center of the Rx antenna. Figure6 shows will make it possible to emulate OAM beam generation. comparisons between a UCA with a 4 cm diameter, a UCA with a 16 cm diameter, and a UCA with a

4 cm diameter plus a 16 cm lens.

Figure 6. Experimental beam diameter results.

Figure 6. Experimental beam diameter results. Next, we conducted experiments to ascertain the beam divergence reduction effect obtained with the lens. We took into account the distance from the center of the Rx side to the location of the Appl. Sci. 2019, 9, x FOR PEER REVIEW 8 of 13

strongest beam intensity since it is the quantitative index for evaluating beam divergence. We defined this metric as the beam diameter. The smaller is the beam diameter, the smaller is the beam divergence at the Rx side because the beam energy is more focused close to the center of the Rx antenna. Figure 6 shows comparisons between a UCA with a 4 cm diameter, a UCA with a 16 cm Appl. Sci. 2019, 9, 1729 8 of 13 diameter, and a UCA with a 4 cm diameter plus a 16 cm lens. The beam diameters of the UCA with a lens in modes 1 and 2 were, respectively, 40 and 60 cm smallerThe beam than diameters that of a ofUCA the UCAhaving with a 4a cm lens diameter in modes without 1 and 2a were, lens. respectively,We also found 40 andthat 60the cm beam smallerdiameter than that of the of aUCA UCA with having a lens a 4 cmwas diameter in good withoutagreement a lens. with We the also beam found diameter that the of beam a UCA diameter with a 16 of thecm UCA diameter. with a These lens was results in good enabled agreement us to confirm with the th beamat the diameter beam divergence of a UCA with can abe 16 reduced cm diameter. by using Thesea dielectric results enabled lens and us tothat conﬁrm the transmission that the beam distan divergencece can be can correspondingly be reduced byusing increased. a dielectric Experimental lens andresults that the also transmission showed that distance a UCA can with be correspondingly a 4 cm diameter increased. and a lens Experimental produced an results OAM alsobeam showed similar to thatthat aUCA produced with a 4by cm a UCA diameter with and a 16 a cm lens diameter. produced Th anese OAM results beam suggested similar that to that using produced a dielectric by a lens UCAis withone of a 16the cm options diameter. that Thesewill enable results OAM suggested wireless that multiplexing using a dielectric systems lens to is effectively one of the address options the thatbeam will enable divergence. OAM wireless multiplexing systems to eﬀectively address the beam divergence.

4. Experimental4. Experimental Demonstration Demonstration Wireless Wireless OAM OAM Multiplexing Multiplexing Technology Technology using using 28 GHz 28 GHz

4.1.4.1. OAM OAM Multiplexing Multiplexing Using Using Multiple Multiple UCAs UCAs TheThe intensity intensity distributions distributions of OAM of OAM signals signals with with diff erentdifferent modes modes are givenare given by the by Besselthe Bessel function function of theof firstthe first kind kind by theirby their inherent inherent nature, nature, as shown as shown in Figure in Figure7. This 7. This yields yields a selective a selective degradation degradation whenwhen the Rxthe antenna Rx antenna is located is located in a low in SNRa low region. SNR Toregion. address To thisaddress problem, this problem, we used multiple we used UCAs, multiple whichUCAs, are designed which are to designed avoid reception to avoid in nullreception regions. in null In this regions. subsection, In this we subsection, first describe we successfulfirst describe experimentalsuccessful results experimental we obtained results using we obtain five OAMed using modes five over OAM 28 modes GHz. over 28 GHz.

Figure 7. Bessel distributions of Rx signals. Figure 7. Bessel distributions of Rx signals. We used multiple UCAs for both Tx and Rx antennas to rectify mode-selective Rx SNR degradation. Figure8 showsWe used our antennamultiple design.UCAs Itfor consists both ofTx fourand UCAs Rx antennas with di ff toerent rectify radii mode-selective and a single antenna Rx SNR in thedegradation. center. Each Figure UCA 8 consists shows our of 16 antenna antenna design. elements. It co Thensists gain of four for each UCAs antenna with different element inradii each and a UCAsingle was 11antenna dBi. The in antennathe center. in theEach center UCA (hereafter consists UCAof 16 No.antenna 0 for elements. notation convenience) The gain for iseach used antenna for the axiselement alignment in each and UCA transmission was 11 dBi. of OAM The antenna mode 0. Wein the used center the following (hereafter two UCA methods No. 0 tofor choose notation differentconvenience) UCAs for is diusedfferent for OAMthe axis modes’ alignment transmission and transmission and chose of Rx OAM UCAs. mode 0. We used the following two methods to choose different UCAs for different OAM modes’ transmission and chose Rx UCAs. •Antenna Selection: Selecting a single Rx UCA that is not located at the null region of each • Antenna Selection: Selecting a single Rx UCA that is not located at the null region of each OAM OAMmode mode •Receiver Diversity: Selecting multiple Rx UCAs to obtain Rx SNR enhancement by • Receiver Diversity: Selecting multiple Rx UCAs to obtain Rx SNR enhancement by receiver receiverdiversity diversity Appl. Sci. 2019, 9, x FOR PEER REVIEW 9 of 13

4.2. OAM-MIMO Multiplexing Using Multiple UCAs This subsection presents wireless OAM-MIMO multiplexing that combines the OAM and MIMO concepts. The number of usable OAM modes is limited due to their mode-dependent beam divergence. In particular, higher OAM modes have a practical limitation due to their nature of large attenuation caused by beam divergence. To address this problem, we present OAM-MIMO multiplexing using multiple UCAs. Unlike OAM multiplexing, OAM-MIMO multiplexing exploits multiple sets of the same OAM modes with receiver equalizations. Consequently, the number of concurrently transmitted data streams can be increased without using higher OAM modes that have large beam divergence. To enable OAM-MIMO multiplexing, we also used multiple UCAs such as OAM multiplexing. To achieve superposition-based simultaneous OAM beam generation and separation, we, Appl.respectively, Sci. 2019, 9, 1729 implemented wideband analog 5 × 16 and 16 × 5 Butler matrices for Tx and Rx9 UCAs. of 13

Figure 8. Implemented multiple uniform circular arrays (Four UCAs and a center antenna). Figure 8. Implemented multiple uniform circular arrays (Four UCAs and a center antenna). 4.2. OAM-MIMO Multiplexing Using Multiple UCAs 4.3.This Experimental subsection presentsOAM Multiplexi wirelessng OAM-MIMO Results Using multiplexing Multiple UCAs that combines the OAM and MIMO concepts.We The conducted number of wireless usable OAM OAM modes multiplexing is limited experiments due to their using mode-dependent five different beam OAM divergence. modes (-2, -1, In particular,0, 1, and 2) higher over 28 OAM GHz. modes Detailed have descriptions a practical re limitationgarding experimental due to their nature parameters of large are attenuation given in Table caused2. by beam divergence. To address this problem, we present OAM-MIMO multiplexing using multiple UCAs. Unlike OAM multiplexing, OAM-MIMO multiplexing exploits multiple sets of the same OAM modes with receiverTable equalizations. 2. Experimental Consequently, parameters (OAM the number multiplexing). of concurrently transmitted data streams can be increased without using higher OAM modes that have large beam divergence. Parameter Value Parameter Value To enable OAM-MIMO multiplexing, we also used multiple UCAs such as OAM multiplexing. Center frequency 28.5 GHz OAM modes -2, -1, 0, 1, 2 To achieve superposition-based simultaneous OAM beam generation and separation, we, respectively, Signal bandwidth 2 GHz Number of streams 5 implemented wideband analog 5 16 and 16 5 Butler matrices for Tx and Rx UCAs. Number of UCAs × 4 × Signal carrier Single carrier 4.3. ExperimentalNumber of OAM antenna Multiplexing Results Using Multiple UCAs 16 Modulation 64 QAM*1 elements in a UCA We conducted wireless OAM multiplexing experiments using ﬁve diﬀerent OAM modes ( 2, 1, Number of antenna − − 0, 1, and 2) over 28 GHz. Detailed descriptions65 regarding experimental Channel coding parameters LDPC are given (DVB-S2) in Table 3/42. elements Frequency domain Diameter of UCATable 2. Experimental 24, 26, 48, 60 parameters cm (OAM Equalization multiplexing). equalization Tx/RxParameter distance 2.5 Value m Parameter Block size Value 256 *1Center T quadrature frequency amplitude modulation, 28.5 GHz *2 low density OAM parity modes check, *3 digital 2,video1, 0,broadcasting 1, 2 − − satelliteSignal bandwidthsecond generation. 2 GHz Number of streams 5 Number of UCAs 4 Signal carrier Single carrier NumberWe ofused antenna the elementssingle carrier with frequency domain equalization (SC-FDE) to average channel 16 Modulation 64 QAM *1 characteristicsin a UCA over a wide signal bandwidth. The signal bandwidth was 2 GHz and 64-QAM modulationNumber of antenna was used elements for all five OAM65 modes. The transmission Channel coding rate per LDPCsingle (DVB-S2) stream was 3/4 9 Gbps. Frequency domain ExperimentsDiameter were of UCA conducted in 24, a shielded 26, 48, 60 room, cm as shown Equalization in Figure 9. The distance between Tx and equalization Rx antennasTx/Rx was distance 2.5 m. The propagation 2.5 m loss can be calculated Block size with Equation (2) with 256 UCA sizes and distance between Tx and Rx. Tx signals were generated by arbitrary waveform generators and fed *1 T quadrature amplitude modulation, *2 low density parity check, *3 digital video broadcasting satellite intosecond a Tx generation. antenna. The Rx antenna received signals and fed them into a digital oscilloscope that

We used the single carrier with frequency domain equalization (SC-FDE) to average channel characteristics over a wide signal bandwidth. The signal bandwidth was 2 GHz and 64-QAM modulation was used for all five OAM modes. The transmission rate per single stream was 9 Gbps. Experiments were conducted in a shielded room, as shown in Figure9. The distance between Tx and Rx antennas was 2.5 m. The propagation loss can be calculated with Equation (2) with UCA sizes and distance between Tx and Rx. Tx signals were generated by arbitrary waveform generators and fed into a Tx antenna. The Rx antenna received signals and fed them into a digital oscilloscope that worked as an analog-to-digital convertor. Since the digital oscilloscope is equipped with four ports, reception was done sequentially using the four ports. This is not significant since the channel environments was static in the shield room. Offline digital signal processing was conducted using all the converted signals. Appl. Sci. 2019, 9, x FOR PEER REVIEW 10 of 13

worked as an analog-to-digital convertor. Since the digital oscilloscope is equipped with four ports, reception was done sequentially using the four ports. This is not significant since the channel environments was static in the shield room. Offline digital signal processing was conducted using all Appl. Sci. 2019the, 9 ,converted 1729 signals. 10 of 13

Figure 9. Experimental environment (OAM multiplexing). Figure 9. Experimental environment (OAM multiplexing). Table3 shows the experimental results we obtained for two methods. For both antenna selection and receiver diversity,Table 3 shows we, respectively, the experimental used results UCA No.we obtain 4, UCAed for No. two 2, methods. UCA No. For 0, both UCA antenna No. 3, selection and UCAand No. receiver 1 for transmitting diversity, OAMwe, respectively, modes 2, used1, 0, +UC1,A and No.+ 24, signals. UCA No. For 2, the UCA antenna No. 0, selection UCA No. 3, and − − method, weUCA selected No. 1 for UCA transmitting No. 3, UCA OAM No. modes 4, UCA -2, No. -1, 0, +1 UCA, and No. +2 signals. 3, and UCA For the No. antenna 1. By applying selection method, the LDPCwe channel selected coding UCA (rate No. 33,/4), UCA we confirmedNo. 4, UCA that No. error-free 0, UCA No. transmissions 3, and UCA were No. obtained1. By applying except the LDPC for the OAMchannel mode coding 2 case. (rate For the 3/4), receiver we confirmed diversity that method, error-free we, respectively, transmissions selected were UCA obtained No. 4except and for the UCA No.OAM 2 for OAMmode modes2 case. For1 andthe receiver+2. We werediversity able method, to confirm we, that respectively, error-free selected transmissions UCA No. were 4 and UCA − received inNo. all 2 OAMfor OAM modes modes with -1 the and same +2. We LDPC were channel able to codingconfirm in that this error-free case. Using transmissions 2 GHz of signal were received bandwidthin yieldedall OAM a modes 45 Gbps with transmission the same LDPC rate. channel coding in this case. Using 2 GHz of signal bandwidth yielded a 45 Gbps transmission rate. Table 3. Experimental results (OAM multiplexing). Table 3. Experimental results (OAM multiplexing). Mode 2 Mode 1 Mode 0 Mode +1 Mode +2 − − Txantenna UCA No. 4 UCA Mode No. -2 2 UCAMode No. -1 0 Mode UCA 0 No. 3 Mode UCA +1 No. Mode1 +2 Antenna Rx antenna UCA No. 3 UCAUCA No. No. 4 UCA UCA No. 0UCA UCA No. 3UCA UCANo. No. 1 Tx antenna UCA No. 1 selection BER (raw) 0.0114 0.02284 0.02012 No. 0.0192 0 3 0.0428 BER (coded) 0.0000 0.0000 0.0000 0.0000 0.0024 Antenna UCA No. UCA No. UCA UCA No. Rx antenna UCA No. 1 selectionTx antenna UCA No. 4 UCA No.3 2 UCA4 No. 0No. UCA 0 No. 33 UCA No. 1 Receiver UCA No. 2, UCA No. 1, Rx antennaBER UCA (raw) No. 3 0.0114 UCA 0.0228 No. 0 0.0201 UCA No. 3 0.0192 0.0428 diversity No. 4 No. 2 BER (raw)BER 0.0105 (coded) 0.0015 0.0000 0.0000 0.0022 0.0000 0.0278 0.0000 0.0385 0.0024 BER (coded) 0.0000UCA 0.0000 No. UCA 0.0000 No. UCA 0.0000 UCA No. 0.0000 Tx antenna UCA No. 1 4 2 No. 0 3 Receiver UCA No. UCA No. UCA UCA No. UCA No. 1, 4.4. Experimental OAM-MIMORx Multiplexing antenna Results Using Multiple UCAs diversity 3 2, No. 4 No. 0 3 No. 2 We conducted experimentsBER on (raw) the OAM-MIMO 0.0105 multiplexing 0.0015 using 0.0022 our implemented 0.0278 antennas. 0.0385 Figure 10 and Table4 show theBER experimental (coded) setup 0.0000 and parameters. 0.0000 Tx 0.0000 signals were 0.0000 generated by 0.0000 offline digital signal processing and fed into our implemented Tx antenna. Outputs of the Rx signals were fed4.4. into Experimental a digital oscilloscope OAM-MIMO that Multiple workedxing as Results an analog-to-digital Using Multiple convertor UCAs as in the OAM multiplexing experiments. In these experiments, the digital oscilloscope was equipped with four ports and receptionWe was conducted done sequentially experiments using on the the ports. OAM-MIMO This also multiplexing did not significantly using our aff implementedect the results antennas. since the channelFigure 10 environments and Table 4 were show static the inexperimental the shield room setup as and in the parameters. OAM multiplexing Tx signals experiments. were generated by offline digital signal processing and fed into our implemented Tx antenna. Outputs of the Rx signals were fed into a digital oscilloscope that worked as an analog-to-digital convertor as in the OAM multiplexing experiments. In these experiments, the digital oscilloscope was equipped with four ports and reception was done sequentially using the ports. This also did not significantly affect the Appl. Sci. 2019, 9, x FOR PEER REVIEW 11 of 13

results since the channel environments were static in the shield room as in the OAM multiplexing Appl.experiments. Sci. 2019, 9, 1729 11 of 13

Appl. Sci. 2019, 9, x FOR PEER REVIEW 11 of 13

results since the channel environments were static in the shield room as in the OAM multiplexing experiments.

Figure 10. Experimental environment (OAM-MIMO multiplexing). Figure 10. Experimental environment (OAM-MIMO multiplexing). Table 4. Experimental parameters (OAM-MIMO multiplexing).

WeParameter used eleven streams for the Value OAM-MIMO mult Parameteriplexing experiment. Two Value Tx UCAs (UCA No. 1 and No. 4), respectively, transmitted five OAM modes (0, ±1, ±2) and UCA No. 0 transmitted OAM Center frequency 28.5 GHz OAM modes 2, 1, 0, 1, 2 − − modeSignal 0. We bandwidth used the received signals 2 GHz of all Rx UC NumberAs for of the streams equalization. Table 11 5 summarizes the modulations,Number of channel UCAs coding rates and 4 corresponding Signal transmission carrier rates used Single in the carrier experiment. The Numberpropagation of antenna loss elementswas calculated with Equation (2) with UCA sizes and distance between Tx and Rx. 16 Modulation 16 QAM/64 QAM We confirmedin a UCA that successful error-free transmissions were obtained in all streams with the usage of LDPC (DVB-S2) Numberthe forward of antenna error elements correction. Total 65transmission rate Channel was 100 coding Gbps at a 10 m transmission distance. 3/4, 5/6, 9/10 The Diameterresults indicated of UCA a new milestone 24, 26, 48, 60was cm reached Equalizationin terms of point-to-point Freq. domain wireless equalization transmission. In addition,Tx/Rx distance we recently extended 10 mour work by extending Block size the baseband signal 256 processing to Figure 10. Experimental environment (OAM-MIMO multiplexing). successfully achieve 120 Gbit/s using a total of 11 OAM modes [19]. This is the state-of-the-art results that have been published in the literature. WeWe used used eleveneleven streams for for the the OAM-MIMO OAM-MIMO mult multiplexingiplexing experiment. experiment. Two Two Tx UCAs Tx UCAs (UCA (UCA No. No.1 and 1 and No. No. 4), respectively, 4), respectively, transmitted transmitted five ﬁveOAM OAM modes modes (0, ±1, (0, ±2)1, and2) UCA and UCANo. 0 No.transmitted 0 transmitted OAM Table 4. Experimental parameters (OAM-MIMO± ± multiplexing). OAMmode mode 0. We 0. used We usedthe received the received signals signals of all of Rx all UC RxAs UCAs for the for equalization. the equalization. Table Table 5 summarizes5 summarizes the themodulations, modulations,Parameter channel channel coding coding rates rates andValue and corresponding corresponding transmission transmissionParameter rates rates used used in the in experiment. theValue experiment. The Thepropagation propagationCenter lossfrequency loss was was calculated calculated with 28.5 with Equation GHz Equation (2) with (2) with UCA OAM UCA sizes modes sizes and and distance distance between -2, between -1, 0,Tx 1, and Tx2 andRx. Rx.We We confirmedSignal conﬁrmed bandwidth that that successful successful error-free error-free 2 GHz transmissions transmissions Numberwere were obtained obtained of streams in inall allstreams streams with with 11the the usage usage of ofthe the forward forwardNumber error error of UCAs correction. correction. Total Total transmission transmission 4 rate rate was was Signal 100 100 Gbps Gbps carrier at at a a 10 10 m m transmission transmission Single carrier distance. distance. TheThe results resultsNumber indicated indicated of antenna a a new new milestone milestone waswas reachedreached inin termsterms ofof point-to-pointpoint-to-point wireless transmission. 16 Modulation 16 QAM / 64 QAM InIn addition, addition,elements we we recentlyin arecently UCA extended extended our work our bywork extending by extending the baseband the baseband signal processing signal toprocessing successfully to achievesuccessfullyNumber 120 Gbit achieve of/ antennas using 120 aGbit/s total using of 11 OAMa total modesof 11 OAM [19]. modes This is [19]. the state-of-the-artThis is the state-of-the-artLDPC results (DVB-S2) that results have 65 Channel coding beenthat publishedhave beenelements inpublished the literature. in the literature. 3/4, 5/6, 9/10 Freq. domain Diameter of UCA 24, 26, 48, 60 cm Equalization TableTable 5. 5.Experimental Experimental resultsresults (OAM-MIMO(OAM-MIMO multiplexing).multiplexing). equalization Tx/Rx distance 10Mode m −2 Mode Block−1 size Mode 0 Mode 1 256 Mode 2 Modulation (QAM) 64 Tx Channel coding rate 3/4 UCA No. 0 Trans. rateTable (Gbps) 5. Experimental results (OAM-MIMO multiplexing). 9 Modulation (QAM) 64 64 16 64 64 Tx Mode -2 Mode -1 Mode 0 Mode 1 Mode 2 ChannelModulation coding (QAM) rate 9/10 3/4 9/1064 9/10 5/6 UCA No.Tx 1 Trans.Channel rate coding (Gbps) rate 10.8 9 7.2 3/4 10.8 10 UCA No. 0 Modulation (QAM) 64 64 16 64 64 Tx Trans. rate (Gbps) 9 ChannelModulation coding (QAM) rate 3/464 3/4 64 9/1016 3/4 64 3/4 64 UCA No.Tx 4 Trans.Channel rate coding (Gbps) rate 9 9/10 9 3/4 7.2 9/10 9 9/10 9 5/6 UCA No. 1 Trans. rate (Gbps) 10.8 9 7.2 10.8 10 5. Conclusions In this paper, we describe how we investigated wireless orbital angular momentum (OAM) multiplexing performance from theory to experimental perspectives. First, we conducted basic theoretical studies on wireless OAM multiplexing performance with respect to various aspects including modulation, demodulation, multiplexing, and demultiplexing. Then, we expanded our interest to experimental evaluation including a dielectric lens and end-to-end wireless transmission at 28 GHz frequency bands. We confirmed that using a dielectric lens can effectively reduce the beam divergence effect and correspondingly increase the transmission distance. In end-to-end experiments, we achieved 45 Gbps throughput using five OAM modes. In addition, we experimentally Appl. Sci. 2019, 9, 1729 12 of 13

5. Conclusions In this paper, we describe how we investigated wireless orbital angular momentum (OAM) multiplexing performance from theory to experimental perspectives. First, we conducted basic theoretical studies on wireless OAM multiplexing performance with respect to various aspects including modulation, demodulation, multiplexing, and demultiplexing. Then, we expanded our interest to experimental evaluation including a dielectric lens and end-to-end wireless transmission at 28 GHz frequency bands. We confirmed that using a dielectric lens can effectively reduce the beam divergence effect and correspondingly increase the transmission distance. In end-to-end experiments, we achieved 45 Gbps throughput using five OAM modes. In addition, we experimentally demonstrated the effectiveness of our proposed OAM-MIMO multiplexing method using a total of 11 OAM modes. In the experiments, we reached a new milestone in point-to-point transmission rates by achieving 100 Gbps at a 10 m transmission distance.

Author Contributions: D.L. and H.S. contributed to conceptualization and methodology, H.S. and H.F. conducted validation, D.L., H.S., H.F. and Y.Y. conducted data curation, D.L. contributed to writing original draft preparation and editing, T.S. supervised overall work. Funding: This research received no external funding. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References

1. Yan, Y.; Xie, G.; Lavery, M.P.J.; Huang, H.; Ahmed, N.; Bao, C.; Ren, Y.; Cao, Y.; Li, L.; Zhao, Z.; et al. High-capacity millimeter-wave communications with orbital angular momentum multiplexing. Nat. Commun. 2014, 5, 4876. [CrossRef][PubMed] 2. Lee, D.; Sasaki, H.; Fukumoto, H.; Hiraga, K.; Nakagawa, T. Orbital Angular Momentum (OAM) Multiplexing: An Enabler of a New Era of wireless Communications. IEICE Trans. Commun. 2017, E100-B, 1044–1063. [CrossRef] 3. Yan, Y.; Li, L.; Zhao, Z.; Xie, G.; Wang, Z.; Ren, Y.; Ahmed, N.; Sajuyigbe, S.; Talwar, S.; Tur, M.; et al. 32-Gbit/s 60-GHz millimeter-wave wireless communication using orbital angular momentum and polarization multiplexing. In Proceedings of the 2016 IEEE International Conference on Communications (ICC), Kuala Lumpur, Malaysia, 22–27 May 2016; pp. 1–6. 4. Lee, D.; Sakdejayont, T.; Sasaki, H.; Fukumoto, H.; Nakagawa, T. Performance evaluation of wireless communications using orbital angular momentum multiplexing. In Proceedings of the 2016 International Symposium on Antennas and Propagation (ISAP), Okinawa, Japan, 24–28 October 2016. 5. Fukumoto, H.; Lee, D.; Sasaki, H.; Kaho, T.; Shiba, H. An experimental study on beam focusing effect using dielectric lens for OAM multiplexing. IEICE Tech. Rep. 2018, 117, 53–57. 6. Lee, D.; Sasaki, H.; Fukumoto, H.; Yagi, Y.; Kaho, T.; Shiba, H.; Shimizu, T. Demonstration of an orbital angular momentum (OAM) multiplexing at 28 GHz. IEICE General Conf. 2018, B-5-90. 7. Lee, D.; Sasaki, H.; Fukumoto, H.; Yagi, Y.; Kaho, T.; Shiba, H.; Shimizu, T. An experimental demonstration of 28 GHz band wireless OAM-MIMO (orbital angular momentum multi-input multi-output) multiplexing. In Proceedings of the IEEE 87th Vehicular Technology Conference (VTC Spring), Porto, Portugal, 3–6 June 2018. 8. Mari, E.; Spinello, F.; Oldoni, M.; Ravanelli, R.A.; Romanato, F.; Giuseppe, F.; Parisi, G. Near-field experimental verification of separation of OAM channels. IEEE Antennas Wirel. Propag. Lett. 2015, 14, 556–558. [CrossRef] 9. Willner, A.E. Communication with a twist. IEEE Spectrum 2016, 53, 34–39. [CrossRef] 10. Mahmouli, F.E.; Walker, S.D. 4Gbps uncompressed video transmission over a 60-GHz orbital angular momentum wireless channel. IEEE Wirel. Commun. Lett. 2013, 2, 223–226. [CrossRef] 11. Cheng, L.; Hong, W.; Hao, Z. Generation of electromagnetic waves with arbitrary orbital angular momentum modes. Scientific Rep. 2014, 4, 4814. [CrossRef][PubMed] 12. Jin, J.; Luo, J.; Zhang, X.; Gao, H.; Li, X.; Pu, M.; Gao, P.; Zhao, Z.; Lou, X. Generation and detection of orbital angular momentum via metasurface. Scientific Rep. 2016, 6, 24286. [CrossRef][PubMed] 13. Deng, C.; Chen, W.; Zhang, Z.; Li, Y.; Feng, Z. Generation of OAM Radio Waves Using Circular Vivaldi Antenna Array. Int. J. Antennas Propag. 2013, 2013, 1–7. [CrossRef] Appl. Sci. 2019, 9, 1729 13 of 13

14. Haskou, A.; Mary, P.; Hélard, M. Error probability on the orbital angular momentum detection. In Proceedings of the 2014 IEEE 25th Annual International Symposium on Personal, Indoor, and Mobile Radio Communication (PIMRC), Washington, DC, USA, 2–5 September 2014; pp. 302–307. 15. Opare, K.A.; Kuang, Y.; Kponyo, J.J.; Nwizege, K.S.; Enzhan, Z. The degree of freedom in wireless line-of-sight OAM multiplexing system using a circular array of receiving antenna. In Proceedings of the 2015 Fifth International Conference on Advanced Computing & Communication Technologies, Rohtak, Haryana, India, 21–22 February 2015. 16. Opare, K.A.; Kuang, Y.; Kponyo, J.J.; Nwizege, K.S.; Enzhan, Z. Mode combination in an ideal wireless OAM-MIMO multiplexing system. IEEE Wirel. Commun. Lett. 2015, 4, 449–452. [CrossRef] 17. Tian, H.; Liu, Z.; Xi, W.; Nie, G.; Liu, L.; Jiang, H. Beam axis detection and alignment for uniform circular array-based orbital angular momentum wireless communication. IET Commun. 2016, 10, 44–49. [CrossRef] 18. Fukumoto, H.; Lee, D.; Sasaki, T.; Nakagawa, T. Beam divergence reduction using dielectric lens for orbital angular momentum based wireless communications. In Proceedings of the 2016 International Symposium on Antennas and Propagation (ISAP), Okinawa, Japan, 24–28 October 2016. 19. Sasaki, H.; Lee, D.; Fukumoto, H.; Yagi, Y.; Kaho, T.; Shiba, H.; Shimizu, T. Experiment on Over-100-Gbps Wireless Transmission with OAM-MIMO Multiplexing System in 28-GHz Band. In Proceedings of the 2018 IEEE Global Communications Conference (GLOBECOM), Abu Dhabi, United Arab Emirates, 9–13 December 2018.

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).