Fair Joint Routing and Scheduling Problem in Wireless Mesh Networks

There is an increasing interest in using Wireless Mesh Networks (WMNs) as next-generation broadband and ubiquituous access network. WMNs is indeed a scalable and cost-effective solution [1]. In such networks, information is routed from source to destination over multiple wireless links, which has potential advantages over traditional single-hop networking, especially for backhaul communication [2]. Industrial standards groups are revisiting the existing protocols and they propose enhanced specifications for WMNs, e.g., IEEE 802.11s and 802.16d. A WMN is usually composed of wireless mesh routers and mesh clients. Wireless mesh routers form a multihop wireless network that serves as the backbone to provide network access for mesh clients. In such networks, wireless mesh routers are usually stationary and connected with a fixed alternating current power, which makes WMNs quite different from the well-studied wireless ad hoc networks and wireless sensor networks. In WMNs, a high volume of traffic is expected to be efficiently delivered on the bandwidth-limited wireless channels, and a large number of users must be fairly served [3]. In such networks, an end-to-end throughput guarantee is an important information to the network operator. In this work, we call gateway the mesh routers that interface with other networks. We simply call router each device that is not a gateway. The routers are up to the job of the access point of the mesh clients. We consider that the traffic goes from all routers to any gateways in the WMN. In wireless networks, the communication channels are shared by the wireless terminals. Thus, one of the major issue is the reduction of capacity due to interference caused by simultaneous transmissions [4]. Multiple channels and multiple radios are just palliatives. Wireless networks performances study motivates various research works. WMNs deployment in operational situations such as urban areas needs QoS guarantees because of capacity constraints. Indeed it has been prooved under some hypotesis that random network performances degrade with a factor at least O(1/ √ n) when its size, n, grows [5, 6, 7]. Some capacity evaluation frameworks