移动自组织网络预测分簇算法的路由协议

Research on Routing Algorithm Based Prediction in Mobile Ad Hoc Networks

针对移动自组织网络中传统分簇算法存在稳定性低、网络开销大的问题,在WCA分簇算法的基础上,提出一种带有预测机制的EWCA-MP(Efficient on-demand Weighted Clustering Algorithm using Mobility Prediction)分簇算法,该算法在簇头选择时充分考虑节点间的链路保持时间,在簇维护阶段引入模糊逻辑的概念,对Hello消息包的广播周期进行优化。并将其应用于CBRP中,提出了一种ECBRP-MP(Efficient Cluster Based Routing Protocol using Mobility Prediction)移动预测的分簇路由协议。仿真结果表明,EWCA-MP算法在簇头数目、单位时间内节点转移次数和统治集更新次数明显减少,ECBRP-MP路由协议在路由开销、分组投递率的性能得到优化。

For the problem that the traditional clustering algorithm in mobile Ad Hoc networks has low stability and high overhead, a prediction mecha- nism clustering algorithm EWCA-MP based on WCA clustering algorithm is proposed. The clustering algorithm in the cluster head node fully considered when selecting between mobility, and in the cluster maintenance phase using the concept of fuzzy logic to optimize the Hello message packet radio cycle. And applied to CBRP, the protocol of efficient cluster based routing protocol using mobility prediction is proposed. The simulation results show that EW- CA-MP reduces the average number of cluster, reaffiliations per unit time and number of dominant set update, and the protocol of ECBRP-MP has pro- moted the performance of routing control overhead and packed delivery ration.