基于遗传算法的道路车辆自组织网络源路由机制

在当前基于交叉路口的城市车辆自组织网络(VANETs)路由协议中，道路上数据包传输大多采用基于地理位置的贪婪转发策略，当数据量较大时，个别节点负载较重，极易引起传输延迟增大乃至丢包的情况.本文提出了一种基于遗传算法的源路由机制，通过记录单体车辆的驾驶信息而非传统方法中的车流均值数据，来预测道路上车辆网络的连通情况，并借助遗传算法，首次基于道路连通性、节点负载和连接跳数这3 点综合考虑，计算得出道路上最佳的源路由节点序列.仿真实验结果表明，在传输率与延迟时间上，性能均优于传统的贪婪路由机制，尤其在车流量为250 veh· lane-1· h-1时，传输率提升约13%.该研究可为智能交通信息通讯提供可靠助力.

An Intra-street Source Routing Mechanism Based on Genetic Algorithm in VANETs

In most of existing intersection-based routing protocols in urban vehicular ad hoc networks (VANETs), geographical greedy forwarding strategy is still adopted for packets delivery in intra- streets. The heavy network load on some certain nodes caused by large data traffic would likely incur large end-to-end delay and even packets dropping. In this paper, we propose an intra-street source routing mechanism based on genetic algorithm (ISSR). By recording the driving data of each individual vehicle instead of the mean value of the traffic flow, we estimate the connectivity in the street. And we are the first to calculate the optimal nodes sequence based on genetic algorithm taking account of the factors of connectivity, node load and hops. The simulation results show that ISSR outperforms the traditional protocol GPSR in terms of packet delivery ratio, average end-to-end delay. In especial, under the condition of 250 veh· lane- 1 · h- 1 , it has a performance improvement about 13% in the packet delivery ratio. This research can provide reliable support for the information communication in ITS.