Traffic-Aware Channel Allocation and Routing in Multichannel, Multi- Radio Wireless Networks.

摘要

Modern day wireless network applications exhibit varying service demands to satisfy user requirements, while differing in the nature of traffic they generate. Future wireless networks should, therefore, be capable of adapting to the heterogeneous traffic characteristics, by e efficiently utilizing the expensive radio resources. In this dissertation, we concentrate on three important problems in existing wireless networks and propose algorithms for addressing them. As the first problem, we focus on the effect of rate-independent MAC overheads in random access protocols such as carrier sensing, backoff , and fixed rate header transmissions. These overheads become prominent in short packets that are transmitted at high data rates. We address this problem by partitioning the transmission spectrum into a narrow channel and a wide channel. The narrow channel is used for transmitting the short packets and the wide channel is used for transmitting the longer packets. We propose a protocol called WiSP (channel Width Selection based on Packet size) to estimate the appropriate channel widths depending on the relative traffic load involving short and long packets in the network. Next, we address the problem of channel switching delay in multichannel, multihop wireless networks. Future networks can benefit from using multiple channels simultaneously within a network. However, to maintain connectivity between various wireless nodes, the wireless radios may have to switch between channels. Due to both software and hardware restrictions switching channels incur significant delay, which can be detrimental to many delay-sensitive, real-time applications, such as VoIP and interactive gaming. To address this, we propose SHORT (Static-Hybrid approach for rOuting Real Time applications), a joint channel allocation and routing algorithm for finding delay efficient routes for real-time applications without significantly a affecting the performance of non-real time applications.