基于无监督聚类的WSN最优路由方法设计

Design for the Method of WSN Optimal Route Based on Unsupervised Clustering

为了解决无线传感器网络中在设计最优路由时无法兼顾提高网络生命周期的问题,提出了一种基于无监督聚类的WSN的最短路由设计方法.首先,提出了基于K-means算法对整个网络进行无监督聚类,并将聚类作为网络的分簇,而聚类中心作为簇头节点;在此基础上,提出了一种基于分簇的簇头到Sink节点的路由算法,寻找出从各簇头到Sink节点的最优路由,当簇头的剩余能量低于阈值时,启动所在簇成员节点中与其距离最短同时剩余能量最高的节点作为新的簇头节点,在NS2仿真环境下,对网络生命周期和数据包传输量进行了仿真.结果表明文中方法较其他方法具有较好的负载均衡性能、网络生命周期以及数据包总传输量.

In order to prolong the life cycle of the network in the optimal route design of wireless sensor network(WSN),a shortest route is designed based on unsupervised clustering.Firstly,the K-means algorithm is provided to cluster the whole network,and the cluster head will be used as the mediate node for route.Meanwhile,the route algorithm for optimizing the cluster head to the sink node is proposed.In order to improve the network life cycle,the cluster member with the biggest remain energy and shortest distance will be regarded as the new cluster head.Therefore,if any node has data packets to send,it only needs to transfer it to its cluster head.The cluster head will send the data packet to the sink node according to the route.In the NS2 simulation environment,the network life cycle and transferred data packets are simulated.The result shows the proposed method has the better load balance performance,network life cycle and transferred data packets.