融合K均值分簇MST路由的无线传感网压缩采样技术

摘要

Considering the special characteristics of data collection and energy constraints of wireless sensor net-works,the paper combines clustered routing strategy with compressed sensing data collection method and then pro-poses a compressed sensing based compressive sampling algorithm with K-Means clustering MST(Minimum Span-ning Tree)routing. The proposed algorithm uses the sparse projection matrix in order to reduce the correlation de-gree value between the projection matrix and sparse matrix so as to reduce the amount of data transmission in the ba-sis to ensure the quality of the data reconstruction by using K-Means clustering MST data fusion tree. The simula-tion results show that this algorithm can improve the network energy usage efficiency,and also be suitable to all kinds of scale wireless sensor networks.%考虑无线传感网中数据采集特点和能量约束性,将分簇路由策略融合到压缩感知采样中,提出了一种融合K均值分簇MST路由的压缩采样算法.算法采用稀疏投影矩阵以减小投影矩阵与稀疏基之间的相关度,利用K均值分簇MST(Mini?mum Spanning Tree)机制构造数据融合树,在保证数据重构质量的基础上减少网络数据传输量.仿真结果表明,算法可以提高网络能量使用效率,同时可以适应各种规模的无线传感网.