蚁群算法结合粒子群算法的WSN路由优化

发布时间：2018-05-05 17:38

本文选题：无线传感器网络 + 改进的蚁群优化算法　；参考：《江南大学》2014年硕士论文

【摘要】：路由环节对于无线传感器网络节省能量非常关键，，因为路由协议决定着节点通信的路径，而通信路径影响着通信的能耗。考虑到层次型路由协议有优于平面型路由协议的性能，本文重点在层次路由协议上开展研究。 (1)利用改进的蚁群优化算法优化WSN簇头多跳最优路径：在DCHS算法分簇的基础上，利用改进的蚁群算法，搜寻从距汇聚节点最近的簇头节点出发，遍历所有簇头节点，最终到达汇聚节点的最优路径。这是一个改进的TSP模型。由于每个簇头之间距离很短，大大减少了每一个簇头都单独与汇聚节点通信的能耗。并且该算法的最优路径是全局最优路径，并不是每个簇头节点都有一条由它至汇聚节点的最优路径。改进的蚁群优化算法体现在，优化了选择概率公式中的启发函数。 (2)利用特殊粒子群算法优化WSN簇头的选取：先用DCHS算法进行预分簇，再采用特殊粒子群算法选择出每个簇内最适合当选簇头的节点。特殊粒子群算法中，每个粒子对应于一个簇，只在簇内跳动，并且不重复之前经过的节点。因此，迭代次数大大地减少。而且由于全局极值因子对每个簇的簇头选取没有参考价值，特殊粒子群算法中没有全局极值因子。它的特点就是迭代次数少，效率高，且能明显延长第一个节点死亡时间。 (3)融合上述两种优化算法，分别作用于簇的建立阶段、簇头与汇聚节点的通信阶段。综合了两种算法的PSO-ACO算法，分别与这两种算法通过仿真图进行性能上的比较。经过验证，PSO-ACO算法很好地均衡了网络能耗，在第一个节点死亡时间、网络的生命周期两方面有了较大的提升。
[Abstract]:Routing is very important for energy saving in wireless sensor networks because the routing protocol determines the path of node communication and the communication path affects the energy consumption of communication. Considering that hierarchical routing protocol has better performance than planar routing protocol, this paper focuses on hierarchical routing protocol. 1) the improved ant colony optimization algorithm is used to optimize the multi-hop optimal path of WSN cluster head. Based on the clustering of DCHS algorithm, the improved ant colony algorithm is used to search the cluster head node nearest to the convergent node and traverse all cluster head nodes. Finally, the optimal path to the convergent node is obtained. This is an improved TSP model. Because of the short distance between each cluster head, the energy consumption of each cluster head communicating with the sink node is greatly reduced. Moreover, the optimal path of the algorithm is the global optimal path, and not every cluster head node has an optimal path from it to the convergence node. The improved ant colony optimization algorithm is embodied in the optimization of the heuristic function in the selection probability formula. (2) using special particle swarm optimization algorithm to optimize the selection of WSN cluster heads: firstly, the DCHS algorithm is used to pre-cluster, then the special particle swarm optimization algorithm is used to select the nodes in each cluster that are most suitable for the selection of cluster heads. In the special particle swarm optimization algorithm, each particle corresponds to one cluster, only beats in the cluster, and does not repeat the nodes passed before. As a result, the number of iterations is greatly reduced. Since the global extremum factor has no reference value for the cluster head selection of each cluster, there is no global extremum factor in the special particle swarm optimization algorithm. It is characterized by less iterations, high efficiency and the ability to prolong the death time of the first node. The fusion of the above two optimization algorithms acts on the establishment of the cluster and the communication between the cluster head and the convergent node respectively. The PSO-ACO algorithm of the two algorithms is synthesized and compared with the two algorithms by simulation graph. It is proved that the PSO-ACO algorithm can balance the network energy consumption well and improve the lifetime of the first node and the lifetime of the network.
【学位授予单位】：江南大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN929.5;TP212.9

【参考文献】