一种结合K-means均匀分簇和数据回归的WSN能量均衡策略

An Energy Balanced Strategy for WSN Combine Uniform Clustering by K-means and Data Regression

针对LEACH协议簇头节点分布不均导致无线传感网节点能量消耗不均衡等不足,提出一种结合K-means均匀分簇和数据回归的能量均衡策略.采用优化初始簇中心K-means算法构建均匀分簇的分级无线传感网,通过获取节点地理位置信息,采用K-means聚类算法形成k个均匀分簇,再选举簇内节点剩余能量最多者当选簇头.该成簇算法可以使网络负载均匀,延长网络生存周期.通过优化初始簇中心的选择,降低K-means算法的迭代次数,使其更快收敛,成簇时间开销更少,簇与簇之间的地理分布也更均匀.在稳定数据传输阶段,采用数据回归的方法来减少普通节点与簇首的通信量,以达到降低功耗的作用.实验结果表明,该策略能够有效降低节点的功耗,延长网络的生存时间.

Aiming at unbalanced WSN（wireless sensor network） energy consumption,a new method was proposed to solve the problem of uneven distribution of cluster head nodes in LEACH protocol. The K-means algorithm of optimize initial cluster center combined data regression is used to structure hierarchical wireless network with uniform clusters. By acquiring the location information of the nodes, the K-means algorithm is intended to form K clusters, and then select the highest residual energy of the nodes in the cluster to be cluster head. The clustering algorithm can balance energy consumption. By optimizing the choice of the initial cluster center, it could reduce iterations of the k-means algorithm, accelerate convergence and cut down the time of clustering and the geographical distribution between the clusters will be more uniform. In the stage of stable data transmission,the data regression could reduce the amount of communication between common nodes and cluster heads in order to reduce power consumption. The strategy can effectively reduce the power consumption of nodes and prolong the lifetime of network as the results of simulation show.