Estimating the channel capacity of multi-hop IEEE 802.11 wireless networks

Abstract

In IEEE 802.11 wireless networks, the residual capacity of the wireless links should be accurately estimated to realize advanced network services such as flow admission control or load balancing. In this paper, we propose an algorithm that estimates the packet delivery failure probability by collecting transmission statistics from nearby nodes, and by using a basic collision detection mechanism. This probability is then used in an analytical model to calculate the maximum allowable traffic needed to reach the saturation condition. We show by simulations that estimation error is within 0.5–5.0%, which is significantly lower than the best performance of prior estimation methods. We also demonstrate that the flow admission control is successfully achieved in a realistic wireless network scenario by the help of accurate link residual bandwidth estimation, where the unsatisfied traffic demand remain bounded at a negligibly low level. A routing algorithm that finds max–min residual bandwidth path between source and destination nodes is also implemented, and simulation results show that the network throughput achieved by this algorithm significantly exceeds that of other popular mesh routing protocols. Finally, we provide test results from the real implementation of our algorithm on 802.11 wireless equipment, which are consistent with the simulations.