Theoretical and experimental study of performance anomaly in multi-rate IEEE802.11ac wireless networks

Abstract

IEEE 802.11 wireless local area networks (WLANs) are shared networks, which use contention-based distributed coordination function (DCF) to share access to wireless medium among numerous wireless stations. The performance of the distributed coordination function mechanism mostly depends on the network load, number of wireless nodes and their data rates. The throughput unfairness, also known as performance anomaly is inherent in the very nature of mixed data rate Wi-Fi networks using the distributed coordination function. This unfairness exhibits itself through the fact that slow clients consume more airtime to transfer a given amount of data, leaving less airtime for fast clients. In this paper, we comprehensively examine the performance anomaly in multi-rate wireless networks using three approaches: experimental measurement, analytical modelling and simulation in Network Simulator v.3 (NS3). The results of our practical experiments benchmarking the throughput of a multi-rate 802.11ac wireless network clearly shows that even the recent wireless standards still suffer from airtime consumption unfairness. It was shown that even a single low-data rate station can decrease the throughput of high-data rate stations by 3–6 times. The simulation and analytical modelling confirm this finding with considerably high accuracy. Most of the theoretical models evaluating performance anomaly in Wi-Fi networks suggest that all stations get the same throughput independently of the used data rate. However, experimental and simulation results have demonstrated that despite a significant performance degradation high-speed stations still outperform stations with lower data rates once the difference between data rates becomes more significant. This is due to the better efficiency of the TCP protocol working over a fast wireless connection. It is also noteworthy that the throughput achieved by a station when it monopolistically uses the wireless media is considerably less than 50 % of its data rate due to significant overheads even in most recent Wi-Fi technologies. Mitigating performance anomaly in mixed-data rate WLANs requires a holistic approach that combines frame aggregation/fragmentation and adaption of data rates, contention window and other link-layer parameters.