针对802.11多射频无线Mesh网络(MR-WMN)不能有效支持端到端低时延组播的问题,首先围绕MAC层传输时延和Mesh层排队时延建模分析,并提出低时延组播路由模型MR-MED(Multi-Radio Multicast End-to-End Delay).其次证明全局流干扰最小化问题是一个NP完全问题且全局流干扰和网络密度的数学关系符合dPlN分布,在此基础上提出有效减小MAC层流内和流间干扰的DCA算法.最后提出流量自适应的组播多径路由方案MMRA,有效减小Mesh层排队时延.仿真与常见算法的比较表明,提出的模型准确刻画了MR-WMN的组播时延,且联合运用DCA和MMRA有效降低了端到端时延.
Multi-radio wireless mesh network(MR-WMN) is one of the key access techniques in Mobile Internet.However,current 802.11-based MR-WMN is not able to provide low multicast end-to-end delay.To address this problem,we first propose a layered and analytical model by combining overlapping channel assignment with multipath routing strategies.The proposed model can be used as a guide on multicast design.It decoupled multicast delay into transmission delay in the MAC layer and queuing delay in the Mesh layer based on that it derived a new multicast routing metric that had low end-to-end delay.Second,we prove that finding the minimum global flow interference solution is a NP-Complete problem and the relationship between global flow interference and network node densities is distributed in accord with double-Pareto lognormal(dPlN) distribution.Based on these two results,the DCA algorithm is proposed in order to minimum global flow interference which can efficiently reduce the multicast transmission delay in MAC layer.Last,to avoid the best wireless link being congested,we propose the flow adaptive-based MMRA algorithm by making use of the MR-MED routing metric and the multipath routing design philosophy,which took local channel congestion into account and can efficiently reduce the queuing delay in Mesh layer.Simulation result and comparison of the common algorithm MCM both show that the proposed model accurately characterizes the multicast delay in multi-radio wireless mesh network and the combination of the DCA and MMRA algorithms efficiently reduce the multicast end-to-end delay.