Performance analysis of self-organised multicast group in multi-radio multi-channel wireless mesh networks
Self-organising is an important characteristics of wireless mesh networks, which provides high flexibility and adaptive connectivity for end users. In this study, we specifically study multicast features in wireless mesh networks to improve network capacity by taking advantage of shared links to reach simultaneously multiple users. Self-organising also provides topology control to manage the dynamic of groups where nodes can join and leave at any time. We evaluate the performance of two algorithms to dynamically create multicast groups. The first one, named Node Joining the Multicast tree, allows nodes to join the multicast tree while reducing the number of relay nodes and minimising interferences by reassignment of wireless channels. The second one, named Node Disjoining the Multicast tree, addresses the multicast tree pruning process when nodes leave the multicast tree. These two algorithms aim to optimise the performance of a multicast tree and to guaranty network connectivity by performing the required channel reassignments. These algorithms also provide an efficient mechanism for recovering from node failures. Using delay and throughput as metrics, simulation results show that the proposed algorithms are able to significantly improve throughput, delay, and robustness compared to existing multicast routing algorithms based on the static shortest multicast tree.