Throughput Gateways-congestion trade-off in Designing Multi-radio Wireless Networks Djohara Benyamina Abdelhakim Hafid Michel Gendreau © Springer Science Abstract In Wireless Mesh Networks (WMNs), traffic is mainly routed by WMN Backbone (WMNB) between the mesh clients and the Internet and goes through mesh gateways. Since almost all traffic has to pass through one of the MGs, the network may be unexpectedly congested at one or more of them, even if every mesh router provides enough throughput capacity. In this paper, we address the problem of congestion of gateways while designing WMNs. We propose a simultaneous optimization of three competing objectives, namely network deployment cost, interference between network channels and congestion of gateways while guaranteeing full coverage for mesh clients. We tailor a nature inspired meta-heuristic algorithm to solve the model whereby, several trade-off solutions are provided to the network planner to choose from. A comparative experimental study with different key parameter settings is conducted to evaluate the performance of the model. Keywords wireless mesh network design problem multi-objective model simultaneous optimization congestion of gateways meta-heuristic method. The research reported in this manuscript has been supported in part by Natural Sciences and Engineering Research Council of Canada (NSERC) and Bell Canada. D. Benyamina ()A. HafidM. Gendreau IRO department, University of Montreal, Montreal, Canada e-mail: benyamid@iro.umontreal.ca A. Hafid e-mail: ahafid@iro.umontreal.ca M. Gendreau e-mail: michel.gendreau@iro.umontreal.ca 1 Introduction The success of the Wireless Mesh Network (WMN) technology has caused a paradigm shift in providing high bandwidth network coverage to users. The Wireless Mesh Network Backbone (WMNB) consists of mesh routers (MRs) interconnected with each other through point-to-point wireless links to provide connectivity to mesh clients (MCs). MRs responsible for providing internet access to clients are called access points (APs) while other more expensive MRs, that are equipped with a gateway capability through which they interface with Internet, are called mesh gateways (MGs). WMNs are highly reliable, scalable, adaptable and cost-effective. They are already pervasive in many diverse environments, such as home networking, enterprises, and universities. Nevertheless, users experience a number of problems, such as intermittent connectivity, poor performance and lack of coverage [1]. In Multi-Radio Multi-Channel (MR-MC) networks, MRs are equipped with multiple network interfaces, thus allowing simultaneous communications over orthogonal channels. However, since the number of available orthogonal channels is limited, interferences happen causing network performance degradation. A proper WMN design is a fundamental task; if addressed carefully it can considerably improve the network efficiency in terms of coverage, throughput, delay and capacity. Basically, the design of WMNs involves deciding where to install the network nodes (given a set of candidate locations), which type of nodes to select (AP, MG or simple MR), how many of these nodes to install, and which channel to assign for each node interface, while guaranteeing users coverage, wireless connectivity and traffic flows at minimum cost. Exploiting the trade-offs among network deployment cost, network throughput, and congestion level of gateways, we propose in this study, a new approach to address the problem of WMNs design. Indeed, minimizing the cost requires stingy resources