An experimental study of inter-cell interference effects on system performance in unplanned wireless LAN deployments

Abstract

In this paper, we report on our experimental study of the effects of inter-cell interference on IEEE 802.11 performance. Due to growing use of wireless LANs (WLANs) in residential areas and settings supporting flash crowds, chaotic unplanned deployments are becoming the norm rather than an exception. Environments in which these WLANs are deployed, have many nearby access points and stations on the same channel, either due to lack of coordination or insufficient available channels. Thus, inter-cell interference is common but not well-understood. According to conventional wisdom, the efficiency of an IEEE 802.11 network is determined by the number of active clients. However, we find that with a typical TCP-dominant workload, cumulative system throughput is characterized by the number of actively interfering access points rather than the number of clients. We verify that due to TCP flow control, the number of backlogged stations in such a network equals twice the number of active access points. Thus, a single access point network proves very robust even with over one hundred clients, while multiple interfering access points lead to a significant increase in collisions that reduces throughput and affects media traffic. Only two congested interfering cells prevent high quality VoIP calls. Based on these findings, we suggest a practical contention window adaptation technique using information on the number of nearby access points rather than clients. We also point out the need for collision-resilient rate adaptation in such a setting. Together these techniques can largely recover the 50% loss in cumulative throughput in a setting with four strongly interfering access points.