Performance evaluation of 802.11 WLAN in a real indoor environment

Performance evaluation of wired networks is a well-known problem and many works consider the performance evaluation of wireless networks as the same problem. Compared to wired networks, wireless networks suffer from the shadow effect and Rayleigh fading which makes its performances difficult to predict. Most of the existing performance stochastic models assume that the radio channel is ideal, and as a result, do not consider the radio environment and the interferences. However, we find that deploying a real wireless LAN (especially indoor WLAN) in different environments leads to quite different performance results. We propose in this paper a method integrating radio characterizations in an analytical performance model. Taking advantage of our model, we are able to predict the performance of a wireless network (throughput, global capacity, etc.) of an indoor 802.11b WLAN, whose radio planning chart is known (through radio planning tool WILDE). We also analyze some special scenarios: network with more than one access point and the hidden terminals problem