Efficient Data Routing for Fusion in Wireless Sensor Networks

摘要

The two most important aspects in wireless sensor networks (WSNs) are the communication cost and detection performance. A lot of research efforts have been taken to improve these aspects. Data routing and centralized or decentralized fusion methods have been studied to help reduce energy consumed in WSNs, since in general WSN is unattended and in hostile areas that could not be reached easily. In this paper, we adopt random walk (RW) as a routing method and use compressive sensing (CS) techniques to recover raw observation data that improves communication cost and detection performance for WSNs. We focus on the design of the projection matrix, in which routing and clustering in WSNs are explicitly involved when CS is applied. Our objective is to have the best data reconstruction, and minimum communication cost and energy consumption.