Cross-layer scheduling for multi-users in cognitive multi-radio mesh networks

Abstract

A wireless mesh network has been popularly researched as a wireless backbone for Internet access. However, the deployment of wireless mesh networks in unlicensed bands of urban areas is challenging because of interference from external users such as residential access points. We have proposed Urban-X, which is a first attempt towards multi-radio cognitive mesh networks in industrial, scientific, and medical bands. Urban-X first controls network topology with a distributed channel assignment to avoid interference in large timescale. In such a topology, we develop a new link-layer transmission-scheduling algorithm together with source rate control as a small-timescale approach, which exploits receiver diversity when receivers of multi-flows can have different channel conditions because of varying interference. For this purpose, mesh nodes probe the channel condition of received mesh nodes using group Request to Send and group Clear to Send. In this study, we establish a mathematical Urban-X model in a cross-layer architecture, adopting a well-known network utility maximization framework. We demonstrate the feasibility of our idea using a simulation on the model. Simulation results show improved network throughput from exploiting receiver diversity and distributed channel assignment under varying external user interference. Copyright (C) 2012 John Wiley & Sons, Ltd.