一种基于自适应代表节点选择的WSN数据收集方法

摘要

采用压缩感知的无线传感网络数据收集方法要求每个节点都参与数据收集,会造成很大的能量浪费.本文提出了一种基于自适应代表节点选择的WSN数据收集方法,在保证压缩感知数据重构精度的同时,减少参与数据收集的节点数.首先,采用主成分分析和混合压缩感知相结合的办法设计稀疏基;然后,通过分析稀疏基的框架势FP(Frame Potential)设计压缩感知的稀疏观测矩阵,从而选择代表节点,以减少参与数据收集的节点数目;最后,根据Sink处数据重构精度,自适应调整稀疏观测矩阵以用作下一时刻数据收集,从而保证数据收集的重构精度.仿真结果表明,该方法有效的降低了网络能耗和数据传输量,同时还保证了每个时刻数据重构的精度.%A shortcoming for the conventional compressive sensing(CS)method in wireless sensor network(WSN)is to require all nodes participating in data gathering process such that the waste of energy inevitably happen.In this paper,a data gathering method based on adaptive representative nodes selection is proposed,which not only ensures the accuracy of data reconstruction,but also reduces the number of nodes involved in data gathering.Firstly,a sparse basic is designed by combining principal component analysis(PCA)and CS;Secondly,according to the frame potential(FP)of sparse basic,a sparse measurement matrix is designed to choose representative nodes for reducing the number of nodes involved in data gathering.Thus,the data reconstruction accuracy in Sink node is guaranteed by adaptive controlling the sparse measurement matrix.Finally,it is shown via simulation results that the proposed method can reduce energy consumption and transmission of network,and ensure the accuracy of data reconstruction.