A Hybrid Data Collection Scheme for Wireless Sensor Networks Using Compressed Sensing

Abstract

In recent years, wide attention has been drawn to how to reduce the energy consumption for data collection in wireless sensor networks (WSNs). Unlike existing energy-efficient data collection models in WSNs, we present a hybrid data collection scheme based on Compressed Sensing (CS) to collect the compressed sensed data in an energy-efficient and load-balanced manner. In our proposed scheme, the data collection process is divided into the intra-cluster data collection process and the inter-cluster data collection process. By modelling the optimal clustering problem as a separable convex optimization problem, we can obtain the optimal clustering size and the optimal data transmission radius. Furthermore, a Cluster Head (CH) chain based on pseudo Hilbert curve is built to collect the compressed sensed data among CHs in an accumulative way. The experimental results show that the proposed method not only reduces the total energy consumption, but also effectively balance the distribution of energy consumption among CHs.