Measurement-driven admission control on wireless backhaul networks

IEEE 802.11 wireless networks perform poorly in the presence of large traffic volumes. Measurements have shown that packet collisions and interference can lead to degraded performance to the extent that users experience unacceptably low throughput, which can ultimately lead to complete network breakdown [12]. An admission control framework that limits network flows can prevent network breakdown and improve the performance of throughput and delay-sensitive multimedia applications. In this paper, we present a measurement-driven admission control scheme that leverages wireless characteristics for intelligent flow control in a static wireless network. Experiments on the 25 node UCSB MeshNet show that the proposed admission control scheme can enhance network performance such that the QoS requirements of real time applications, such as VoIP, can be met.