Distributed Compressed Sensing In Wireless Sensor Networks

Wireless sensor networks (WSN) are composed by a large number of wireless sensor nodes, It is responsible for sensing the information of storage, transmission, processing and other functions in a region. Due to the large number of WSN nodes with the huge amount of collected data, the limited energy, it is an urgent problem that how to reduce the power consumption by using this unique nature of WSN to compress data.In recent years, the birth of a new signal processing technique--compressed sensing (CS) theory. Because of its low encoding complexity, less the value of the desired signal sampled, independent of each other codes and other characteristics, efficient data compression and simple encoding algorithm is just meet the application requirements and other aspects of WSN in resource-constrained. CS on the basis of theory, people between WSN further explores the relevance of each node, a new group of information based signal processing technique of multiple signals——Distributed Compressed Sensing (DCS) theory, it reduce the sampling rate of the signal, and the amount of information transmitted over the network through by using signal correlation with each other now closed to further, So Distributed Compressed Sensing technology applications in wireless sensor networks will be very broad.Firstly, this paper introduces the concept and application of WSN first, in view of its network characteristics:energy limited computing power and low limiting factor proposed CS based information processing technology. the basic framework of CS is made a profound exposition and discussion, especially for sparse signal representation, important technical observation matrix structure and signal reconstruction algorithm design.Secondly, based on the CS theory, the DCS intensive study of the theory is evolved with the three Joint Sparse Model(JSM) of DCS frame detail. Various recovery algorithms are analyzed and compared under the JSM-2model.Finally, an independent model is established based on CS-OMP algorithm and joint reconstruction model DCS-SOMP reconstruction algorithms for wireless sensor networks, CS/DCS reconstruction algorithms are signal processing is applied to the group’s perception. Simulation results show that the same effect to reconstruct, DCS/CS reconstruction algorithm requires fewer number of observations compared with the independent joint reconstruction algorithm, which can further reduce energy consumption in wireless sensor networks. It can be concluded that the DCS algorithm based on JSM has outstanding advantages in signal reconstruction performance of WSN.