Research On QoS Multicast Routing Metric In Wireless Mesh Networks

With the sharp development of information industries, and the rapidadvancement of the triple-play process of communications networks, cableTV networks and IP networks, the multimedia services are becoming themain businesses in the networks. Since the multimedia services transmit alarge volume of data, the matter of how to effectively utilize the channelbandwidth has become an urgent problem for the operators. Compared to theresources wasting in traditional communication mode, the multicasttechnology has just solved this problem. At the same time, users desire amanner to ensure the QoS (Quality of Service). In order to guarantee theQoS, high-performance routing is indispensable. As the routing criterion,routing metric has a far-reaching influence on network performance.Therefore, in order to enhance the routing performance and ensure theQoS, this paper, based on wireless mesh networks which are the mainsupport of the access networks for the next-generation Internet, throughstudying about the routing metrics, improves the QoS constraints multicastrouting metric function.In first segment (chapter II), it compares the typical routing metrics ofwireless mesh networks. HOP (Hop Count) considers the “length” of thepath. ETX (Expected Transmission Count) considers packet loss rate of thelink. On the basis of ETX, ETT (Expected Transmission Time) routes a pathwhich has the shortest transmission time. And based on the ETT, WCETT(Weighted Cumulative Expected Transmission Time) captures the intra-flowinterference. Also based on the ETT, INX captures the inter-flowinterference. The simulation results show that the performance of INX isbetter than others in networks throughput, but INX does not take intoaccount the intra-flow interference, and with the increasing of networksload, the packet loss rate increased heavily.Based on the above studies, the third chapter in the paper, throughanalyzing the consequences and causes about intra-flow and inter-flowinterference, proposes a routing metric FIA (Flow Interference Aware)which is comprehensively aware of flow interference. The metric takeschannel as unit and captures the transmission data sizes of all interfered link of each hop in path to measure the inter-flow interference. Furthermore,the metric measures the usage of each channel by getting the maximum ofthe interfered transmission data sizes to consider both of intra-flow andinter-flow interference at overall. Finally, we sum up the measurementresult and the expected transmission time of the path with a weight forbalanced to consider both of the transmission delay and flow interference.Simulation results show that the FIA performs better than the others interms of throughput and packet loss rate.Finally, summarized the research on routing metrics, the chapter IVimproves the QoS constraints multicast routing metric function, andproposes a MQCMR-Ant (Multi-QoS Constraints Multicast Routing basedon Ant Colony Algorithm). By adopting the reciprocal of the product offorward delivery ratio and reverse delivery ratio, MQCMR-Ant converts thepacket loss rate from multiplicative parameter into additive parameter. So itcould effectively measure the packet loss from the side with simplifying thealgorithm. At the same time, MQCMR-Ant transforms the minimum channelbandwidth constraint into the ratio between transmission rate of interferedneighbor data streams and channel bandwidth. And it configures theadjacent hop in the path using different channel to transmission. So, withcapturing the inter-flow interference, intra-flow interference and channelbandwidth, MQCMR-Ant carries out a more practical and comprehensiveconsideration to the path throughput. Simulation results show thatMQCMR-Ant has good expansibility, low run-time, high availablebandwidth, less accounted for network resources, and the constructedmulticast tree is superior to the traditional routing algorithm——ACQMR.