Analysis of Coexistence between IEEE 802.15.4, BLE and IEEE 802.11 in the 2.4 GHz ISM Band Radhakrishnan Natarajan∗y, Pouria Zand∗, Majid Nabiy ∗Holst Centre / IMEC-NL, High Tech Campus 31, 5656 AE Eindhoven, The Netherlands yDepartment of Electrical Engineering, Eindhoven University of Technology, 5600 MB Eindhoven, The Netherlands Email: ∗[email protected], ∗[email protected], [email protected] Abstract—The rapid growth of the Internet-of-Things (IoT) 2405 MHz2405 2410MHz 2415MHz MHz2420 MHz2425 MHz2430 2435MHz MHz2440 MHz2445 MHz2450 2455MHz 2460MHz MHz2465 2470MHz MHz2475 has led to a proliferation of low-power wireless technologies. 2480MHz Frequency 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 A major challenge in designing an IoT network is to achieve 11 Ch coexistence between different wireless technologies sharing the unlicensed 2.4 GHz ISM spectrum. Although there is significant literature on coexistence between IEEE 802.15.4 and IEEE IEEE 802.11 IEEE 802.11 IEEE 802.11 802.11, the coexistence of Bluetooth Low Energy (BLE) with other Ch 1 Ch 6 Ch 11 2 0 1 3 4 5 6 7 8 9 22 33 37 10 38 11 12 13 14 15 16 17 18 19 20 21 23 24 25 26 27 28 29 30 31 32 34 35 36 39 technologies remains understudied. In this work, we examine Ch coexistence between IEEE 802.15.4, BLE and IEEE 802.11, which are widely used in residential and industrial wireless applications. 2408 MHz 2408 MHz 2450 MHz 2472 2402 MHz 2402 MHz 2404 MHz 2406 MHz 2410 MHz 2412 MHz 2414 MHz 2416 MHz 2418 MHz 2420 MHz 2422 MHz 2424 MHz 2426 MHz 2428 MHz 2430 MHz 2432 MHz 2434 MHz 2436 MHz 2438 MHz 2440 MHz 2442 MHz 2444 MHz 2446 MHz 2448 MHz 2452 MHz 2454 MHz 2456 MHz 2458 MHz 2460 MHz 2462 MHz 2464 MHz 2466 MHz 2468 MHz 2470 MHz 2474 MHz 2476 MHz 2478 MHz 2480 We perform a mathematical analysis of the effect of cross- Frequency BLE 802.15.4 technology interference on the reliability of the affected wireless network in the physical (PHY) layer. We also set up and perform Fig. 1. 2.4 GHz ISM band: IEEE 802.15.4, BLE and IEEE 802.11 channels PHY layer experiments to verify the analytical results. Finally, we extend the study to the Medium Access Control (MAC) layer. Our findings show that, even though the MAC layer mechanisms of works present as interference and performs interference miti- IEEE 802.15.4 and BLE improve reliability, cooperative solutions are required to achieve coexistence. gation. In cooperative coexistence, all the networks collaborate and coordinate their use of the spectrum in a fair way. I. INTRODUCTION An example use case of cooperative coexistence is a smart gateway that coordinates spectrum usage between multiple In recent years, the emergence of IoT has led to an networks that employ different wireless technologies. explosive growth in the number of smart wireless devices. In order to propose a coexistence solution for a given Many IoT applications have their own Quality-of-Service application, it is first necessary to study the effect of cross- (QoS) requirements. In response, various low-power wireless technology interference on the performance of the involved technologies such as IEEE 802.15.4 [1] and BLE [2] have wireless technologies. While existing studies in this topic area been developed and adopted widely. The globally allowable provide an insight into the nature of the coexistence issue, they operational band for these technologies is the unlicensed 2.4 fail to systematically analyse the effect of cross-technology GHz ISM band as shown in Fig. 1, which is also used by interference. As a result, the outcomes are quite inconclusive other wireless technologies such as IEEE 802.11 [3]. The and sometimes even contradictory. Furthermore, no research result is cross-technology interference that affects the QoS of on coexistence between IEEE 802.15.4, BLE and IEEE 802.11 the network, particularly its reliability and latency, which may has yet been undertaken, even though these technologies are in turn lead to application failure. Achieving coexistence of widely used in smart homes, smart buildings and industrial different wireless technologies is therefore a major challenge wireless applications. in designing an IoT network. To try and fill this knowledge gap, we perform a systematic Coexistence between different wireless technologies can analysis of coexistence between IEEE 802.15.4, BLE and be classified into three domains: space, time and frequency. IEEE 802.11b. The choice of IEEE 802.11b is arbitrary; this Coexistence can thus be achieved by meeting one or more of study can also be applied to other variants of IEEE 802.11. the following conditions: The contributions of this work can be summarised as follows: 1) Adequate spacing between the networks 1) Adopting the coexistence methodology introduced in [4] 2) Controlled time-sharing of the channel and applying it to IEEE 802.15.4, BLE and IEEE 802.11b 3) Adequate frequency separation between the networks 2) Performing a mathematical analysis to study PHY layer co- Coexistence mechanisms can be classified into two cat- existence of these technologies and verifying the analytical egories: non-cooperative and cooperative, depending on results through experiments whether the involved networks operate independently of one 3) Studying MAC layer coexistence through experiments another or coordinate their use of the spectrum. In non- 4) Providing a basis to connect this study to real-world cooperative coexistence, each network treats the other net- application requirements The remainder of this paper is organised as follows. Section II discusses some of the existing literature on coexistence. (0, D) Section III describes the mathematical analysis. Section IV dis- cusses the outcomes of our experiments. Section V discusses the findings of our study. Section VI presents our conclusions. Finally, Section VII discusses the premises of this study as well as the scope for future work. D Network II. RELATED WORK Affected Wireless Affected There is a significant amount of literature available on co- existence between IEEE 802.15.4 and IEEE 802.11. Likewise, Interfering Wireless Network the coexistence of Bluetooth Classic [5] with other technolo- (0, 0) (d, 0) (e, 0) gies has been studied extensively. However, the coexistence of d BLE with other technologies remains understudied. IEEE 802.15.2-2003 [6] specified recommended practices Fig. 2. Geometric model from [4] for coexistence of IEEE 802.15 Wireless Personal Area Net- works (WPANs) with other networks operating in the same unlicensed frequency bands. The IEEE 802.19 Wireless Coex- III. MATHEMATICAL ANALYSIS istence Working Group has extended this scope by developing In our mathematical analysis of coexistence, we adopt the standards for coexistence between wireless technologies in the methodology developed in [4] to estimate the PER caused by unlicensed frequency bands. Their contributions to the field cross-technology interference, and apply it to IEEE 802.15.4, included a coexistence methodology [4], which analytically BLE and IEEE 802.11b. This PER estimate can be used estimates the effect of cross-technology interference on the to estimate other performance metrics such as latency and reliability of networks. throughput. To make this paper self-contained and to ensure R. G. Garroppo et al. [7] researched the effect of IEEE continuity, we include the derivation of the generic method- 802.11 and Bluetooth Classic interference on IEEE 802.15.4 ology from [4] here, and discuss how we apply it to these and vice versa through experiments. They found that the technologies. In this analysis, we consider only the PHY layers Packet Error Rate (PER) of IEEE 802.15.4 drops by around of the involved wireless technologies. The effect of MAC layer 40% due to IEEE 802.11 interference and by less than mechanisms is considered in the experimental study discussed 10% due to Bluetooth Classic interference. Moreover, IEEE in Section IV. 802.15.4 is more affected by the distance to the IEEE 802.11 A. Methodology interferer than Bluetooth Classic. They also observed negli- gible effects on IEEE 802.11 or Bluetooth Classic, due to The methodology developed in [4] takes the geometric interference from IEEE 802.15.4. model of the Affected Wireless Network (AWN) and the Inter- S. Silva et al. [8] studied the effect of IEEE 802.11, fering Wireless Network (IWN) as its starting point. Following IEEE 802.15.4 and Bluetooth Classic interference on BLE this, a path-loss model is used to calculate the average Signal- through experiments. They observed no effect on the PER to-Interference Ratio (SIR) at the AWN receiver, reflecting the and the Received Signal Strength Indicator (RSSI) of BLE signal and interference transmit powers and the geometry of as a result of interference from any of the other technologies. the networks. A PHY layer model is then used to calculate the This indicates that frequency hopping in BLE is very effective Symbol Error Rate (SER) of the AWN as a function of the in interference-avoidance. SIR at the AWN receiver, assuming continuous interference. J. Wyffels et al. [9] researched the interference effect of BLE Finally, a temporal model takes into account the dynamic advertising beacons on IEEE 802.11 through experiments. nature of the interference by modelling it as a pulse generator They observed that the impact of interference is significantly with known statistical properties, and calculates the AWN’s shaped by channel separation. At a channel separation of 70 PER as a function of SER. MHz, they observed practically no impact on IEEE 802.11 We analyse the following four network configurations: traffic, whereas at 1 MHz, they observed a drop of around 1) AWN = IEEE 802.15.4, IWN = BLE 50% in IEEE 802.11 throughput.