An experimental analysis for compressive sensing in wireless sensor networks

We design a kind of wireless sensor network node for collecting parameters about environments of farmland, and employ compressive sensing algorithm in wireless sensor networks. Temperature is picked out for compressing and reconstructing. Primary analyses show that the algorithm can effectively compress observations for transmitting fewer packets. The increase in computation brought by compressive sensing is within acceptable range, and the total energy cost of the network is reduced by a large ratio. By taking comparison between original and recovery, the results quantitatively show that, compressive sensing algorithm possesses significant performance and efficiency, and can meet requirements of accuracy on signal compression in some applications of wireless sensor network.