基于拓扑连通概率的车载自组织网络路由算法

Routing algorithm based on probability of topology connectivity in vehicular ad hoc networks

为了改进REAR算法中竞争延迟函数存在的不足,减少广播报文数量,从3个方面对其进行了改进,提出了RPR算法.首先,修改了竞争延迟函数的参数,通过节点之间的数据交互,计算得到节点的排名信息,并将此信息作为竞争延迟函数的参数;其次,减少广播报文的传播时间以及数据报文的广播次数,达到抑制报文数量的目的;最后,通过移除REAR算法中节点对于上一跳节点的概率累加,简化了下一跳节点的判断依据,增加了算法的执行力.在NS2仿真软件中实现了REAR算法和RPR算法.对比实验结果表明,RPR算法可将广播报文数量和数据通信时间分别缩短至REAR算法的78%和72%,覆盖率则从原来的93%提升至99%.

To overcome the shortcomings of the contention delay function in REAR（receipt estimation alarm routing） algorithm and reduce the number of broadcasting packages,improvement is put forward in three aspects and a new algorithm called RPR（receipt-based probabilistic routing） algorithm is proposed.First,the parameters of the contention delay function are modified.By exchanging data among the nodes,the ranking information can be obtained and used as the parameter of the contention delay function.Secondly,by shortening the transition time of broadcasting packages and reducing the number of data package transition,the number of packages is reduced.Finally,by cutting back the probability accumulation from previous nodes,the conditions of judging next-hop nodes are simplified and the performance of the algorithm is improved.Both the REAR algorithm and the RPR algorithm are implemented in NS2.The contrast experiment results show that the number of the broadcasting packages and the transition time in the RPR algorithm can be reduced to 78% and 72% of those in the REAR algorithm,and the coverage rate can be increased from 93% to 99%.