DTN中基于动态多维度组合的拓扑图路由算法

摘要

为提高延迟容忍网络中数据包的传递命中率、减少网络传输延时和网络开销,提出一种基于动态多维度组合的拓扑图路由算法.每个节点动态地选择其他一些节点,这些节点与该节点在多个维度的每一个维度中的排名都在前s等级,通过这种方式每个节点构成自身相对静态的局部拓扑图.随着时间的推移,动态选择体现在每个节点所选择的构成自身局部拓扑图的节点集合会发生变化.根据节点间的不断相遇扩大节点自身的拓扑图.通过弗洛伊德最短路径算法得出每个节点的最佳传递集合进行数据包的转发.仿真结果表明,与Prophet路由算法、Bubble Rap路由算法以及SimBet路由算法相比,该算法在平均传输时延、传递命中率、网络开销方面得到较大改善.%In order to improve the delivery hit ratio and reduce network transmission delay and network overhead of data packet in Delay Tolerant Network (DTN),the Topological Graph routing algorithm based on Dynamic Multidimensional Combination(TGDMC)is proposed.Each node dynamically selects nodes that rank in the top s level with the node itself in multiple dimensions to form the relative static local topological graph.As time goes on,the dynamic selection is reflected in the change of the node sets that constitute the local topological graph of each node.According to the continuous meeting between nodes,each node expands its topological graph.Floyd shortest path algorithm is used to obtain the best transfer node set of each node to forward the data packet.Simulation results show that,compared with the Prophet routing algorithm,Bubble Rap routing algorithm and SimBet routing algorithm,the proposed algorithm has a better improvement on average transmission delay,delivery hit ratio and network overhead.