Wireless Contention in Mobile Multi-hop Networks

Wireless contention is the most important and least studied characteristic of multi-hop networks. It is the most important because not only does it have a drastic effect on the performance but also it is the fundamental property which makes these networks very different from traditional wired and single-hop wireless networks. Finite bandwidth is only one of the sources of contention. In addition, there is extensive interference among nearby links irrespective of whether they have any common nodes. This is usually alleviated by employing a scheduling mechanism. Moreover, even transmissions from quite distant nodes outside the scheduling area may interfere with a link due to multi-path fading. Contention is one of the least studied properties because of its complicated nature. This forces the researchers to use simple, intuitive, but often unrealistic models to keep the analysis tractable, which may lead to significantly inaccurate conclusions. This paper introduces a mathematical framework to analyze wireless contention that models all three manifestations of contention, namely, finite bandwidth, local scheduling, and interference from transmissions outside the scheduling area. The framework can be used with any realistic channel and mobility model, allows the derivation of exact, rather than asymptotic, results, and it is quite accurate. As a case study, we use the framework to compute the expected packet delay of two popular routing schemes for mobile ad hoc networks.