无线传感器网络本地集中式动态成簇算法研究

Study of local centralized dynamic clustering algorithm in WSN

为了缩短簇型无线传感器网络的成簇时间、延长网络寿命,提出了一种本地集中式动态成簇算法。对LEACH算法的成簇过程进行了认知与分解,指出了簇密度不均匀是成簇时间延长的主要原因。在此基础上建立了总体设计模型,对簇密度进行了自适应调整,采用簇内节点数阈值控制的簇分裂和簇合并,从而避免了簇密度过大或过小的情况。提出了簇头意外失效时补选簇头和网络层功率控制优化策略。改进的J-Sim仿真实验结果表明,该算法有效地延长了网络寿命。

To curtail wireless sensor network clustering time and extend the network lifetime,an algorithm of local centralized dynamic clustering is presented.Firstly,the process of clustering in LEACH is recognized and decomposed.It is pointed out that the prime reason in the delay of clustering time is the asymmetry of cluster density,and the general design model is constructed.Then,the cluster density are adjusted adaptively by cluster split and mergence according to the threshold of intra-cluster nodes number so as to prevent the cluster density that is not too high or too low.By-election of cluster head is proposed to handle the accidental nodes hardware failures.The optimization of power control in the network level is also proposed.Finally,by improved J-Sim,the function of prolonging network lifetime of the presented method is demonstrated.