The Key Technologies Of Distributed Compressed Sensing In Wireless Sensor Networks

Wireless Sensor Networks(WSN) has a large number of nodes which are composed with the same structure, and its resources are constricted. Therefore, it is a problem to be solved that how to meat the application requirement for WSN in an energy efficient way by compressing the sensing data of nodes through the correlation of data.In recent years, a new signal compression and encoding theory, which is called Compressed Sensing(CS), is gradually developed. It has low complexity of encoding and excellent compression performance, and its encoding and decoding is independent. These advantages make it well suitable for the WSN whose resources are constricted. Furthermore, the sensing data of WSN node has spatial and temporal correlation, and the concentration of Distributed Compressed Sensing(DCS) is to reconstruct more signals by using the intra-correlation and inter-correlation of signal. Therefore, DCS has a broad prospect of application in WSN.At first, we introduced the research situation and application of wireless sensor networks in detail in this paper, and we also had a more profound elaboration and discussion of the basic theoretical framework of compressed sensing, especially for the key technologies such as the sparse transform of signal, the compression and measurement of signal, the reconstruction of signal, and so on.Then, according to the spatial correlation of sensing data of wireless sensor network nodes, spatial correlation-based distributed compressed sensing model was established. The encoding and decoding algorithms were also presented on the basis of this model. In addition, the relation between reconstruction error and compression ratio of distributed compressed sensing was been studied, too.On the basis of the above, we studied the spatial and temporal correlation of the sensing data and the jointly sparse models, and we established the distributed compressed sensing model based on the jointly sparse models and the spatial and temporal correlation in the paper, as well as the scheme of encoding and decoding corresponding to the model. We also carried out a detailed comparison between the joint reconstruction algorithm and the independent reconstruction algorithm of distributed compressed sensing.At last, we verified the models and algorithms proposed in the paper by the simulation. The simulation results show that the distributed compressed sensing algorithm can significantly reduce the data compression ratio, which means that the model has a higher storage efficiency. Therefore, we can conclude that: the distributed compressed sensing algorithm based on JSM2 and spatial and temporal correlation in wireless sensor networks has certain advantages of reconstruction and compression performance, and it will bring to good feasibility for the practical application of wireless sensor networks.