无线传感器网络中基于小波的分布式存储

Distributed Storage for Wireless Sensor Networks Based on Wavelet Hierarchy

无线传感器网络主要用于从目标对象收集信息,由于其能源极其有限,分布式数据存储和查询得到越来越多人的注意.本文提出了一种基于小波构架的新型分布式存储方式,它使所有信息经小波压缩后平均分布于各个节点之中,构成小波系数空间存储结构树.通过仿真实验,表明这种算法在无线传感器网络的数据管理中获得了良好的效果:(1)通过简化小波变换消除了额外的计算和通信量,大大节省了数据管理所需的能耗;(2)利用传感器节点内和节点间的信息关联,有效提高了存储效率;(3)利用小波多分辨率的编码技术和小波系数空间结构树的自相似性,支持时空两个方向的快速查询.

The Wireless Sensor Network （WSN） is mainly used to gather data information from the monitored environments. However, due to the strict constrains on individual sensor node＇s resource, distributed storage and query must be taken. This paper proposed a novel distributed approach based on wavelet hierarchy, which uses wavelet compression techniques to construct a Spatial Architecture Tree and distribute the sensing data in various nodes. The simulation results showed that this approach perform well in database management： a） simplify the lifting scheme of traditional wavelet to eliminates extra transmission and computation, reduce energy costs significantly; b） exploit the inherent correlations that exist in or between sensor readings to store data efficiently; c） support highly efficient spatio-temporal query utilizing wavelet multi-resolution analysis and spatial self-similarity of Spatial Architecture Tree.