Traffic Scheduling with Efficient Channel Assignment in WLAN Mesh Networks

Abstract

Wireless mesh networks have received increasing interest in recent years due to the fast expanding Wi-Fi market. More and more, individual communities and companies are beginning to set up intranets with multiple Wi-Fi access points, so that clients can communicate using wireless connections. Because the traditional IEEE 802.11 standard cannot provide efficient performance for mesh networks, the IEEE set up a task group in 2001 to establish a wireless mesh network standard, IEEE 802.11s. For compatibility, IEEE 802.11s will be an extension of the IEEE 802.11 MAC/PHY, and as a result, the new standard has inherited both the pros and cons of IEEE 802.11. Co-channel and inner-channel interference are the dominant factors affecting the system performance of wireless networks. Since there are a number of available non-overlapping channels one can always use these in order to eliminate inner-channel interference. However, the number of channels is not sufficient for an ESS Mesh if channel reuse is not considered, not only because of the network size, but also because of the non-licensed nature of the IEEE 802.11 PHY ISM band, where the network will suffer interference from other co-located networks. For this reason, channel reuse in an ESS Mesh is essential and reducing co-channel interference is a key issue in channel assignment. In this thesis, we investigate the performance of deterministic traffic scheduling with channel assignment in an ESS Mesh based on a TDMA MAC framework while still using the IEEE 802.11 PHY. We first analyze an upper bound on channel assignment performance, considering both binary interference models and cumulative interference models. Then, a scheduling solution for deterministic traffic is proposed, based on heuristic channel assignment and path selection algorithms. Our simulation results show that the scheduling solution is feasible and the performance is close to the theoretical value.