考虑到无线传感器网络(WSN)负载不均衡导致节点存活时间较短、能量消耗量较多的问题,提出一种基于分布式层次化结构的非均匀聚类负载均衡算法(DCWSN)。首先,建立了一个WSN的多层分簇的网络拓扑结构,并分析了该网络拓扑的簇内节点运作的能量消耗方式。接着,采用非均匀聚类的负载均衡算法,在簇头的选择上考虑了节点连通密度、节点剩余能量和簇头选择时间,通过竞选出最高权重的节点成为簇头;在簇的建立阶段,通过簇大小的决定阈值和簇头的更新机制来均衡簇头的能量负载,防止簇头节点过早死亡。通过网络生命周期和网络能量消耗对提出算法的有效性进行验证,并与算法EDDIE、M-TRAC、DDC和EELBC进行比较,结果显示DCWSN算法的节点存活率为37.7%,高于对比算法,且能量效率也高于对比算法。实验结果表明,DCWSN算法对节点负载分配具有良好的均衡性,有效控制了节点负载过量的问题,提高了节点的能量效率。
Aiming at the problems such as short survival time and great energy consumption caused by load imbalance in Wireless Sensor Network ( WSN), a kind of load balancing algorithm for non-uniform clustering with distributed hierarchical structure named DCWSN was proposed. First, the network topology structure with muhilayer clusters for WSN was established, and the energy consumption mode of the nodes in the clusters of it was analyzed. Second, the load balancing algorithm for non-uniform clustering was used to choose the node with highest weight to be cluster head node with considering the node connected density, node residual energy and cluster head choose time. In the establishment of the cluster stage, the energy load of the cluster head was balanced by cluster size decision threshold and the cluster updating mechanism to prevent the premature death of cluster nodes. Comparison experiments on life cycle of the network and the network energy consumption were conducted with EDDIE, M-TRAC, DDC and EELBC to verify the effectiveness of the proposed algorithm, and DCWSN achieved a higher survival rate of node at 37.7% and a higher energy efficiency. The experimental results show that DCWSN has good performance in load balance, effectively controls the overload of node, and also improves the energy efficiency of node.