无线传感网中一种基于压缩感知的数据存储机制  

A DATA STORAGE MECHANISM BASED ON COMPRESSIVE SENSING IN WIRELESS SENSOR NETWORKS

作　　者：庞慧[1] 高丽婷[1] 孟凡兴[1] 王庆林[1] 

机构地区：[1]河北建筑工程学院计算机系,河北张家口075000

出　　处：《计算机应用与软件》2014年第4期111-114,144,共5页Computer Applications and Software

基　　金：国家自然科学基金项目(61271125/F010 909)

摘　　要：能量有效地实现分布式数据存储是无线传感网中的一个研究热点。针对现有的存储算法的不足,提出一种基于压缩感知的分布式数据存储机制。首先对于各个传感器节点的数据基于压缩感知进行测量;然后从网络中随机选择Ns个节点作为源节点以概率p向其邻居广播自身的测量值,使得数据收集者只需要访问部分节点就能收集整个网络的测量值;最后在保证数据重构精度的前提下,通过优化测量矩阵和转发概率p,实现数据的高效存储。理论分析和仿真实验结果表明,该方法是有效的,在数据转发次数以及能量有效性等方面要优于传统的方法。To realise distributed data storage with energy efficiency is a focus in wireless sensor networks research. Aiming at the disadvantages of existing storage algorithms,we propose a compressive sensing-based new data storage mechanism. First,it measures the data of each sensor node based on compressive sensing technology,then it selects Nsnodes randomly from the network as the source nodes,and broadcasts to their neighbours their own measurements with probability p,which makes the data collectors can gather the measurements of whole network by only accessing to a small number of nodes. Finally,under the premise of ensuring data reconstruction precision,the efficient data storage is realised through the optimisation of measurement matrix and forwarding probability p. Theoretical analysis and simulation experimental results show that,our method is effective. It is better than the traditional methods in times of data transmission and energy efficiency.

关 键 词：无线传感网 数据存储 压缩感知 测量值 数据转发次数 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]