A Method of Signal Sparse in Wireless structural health monitoring based on Compressive Sensing

Compressive sensing (CS) is a novel compressive sampling method applied to the compressible signal in wireless sensor networks (WSN) based structural health monitoring. Using this method, the number of sensors for SHM is greatly reduced, so that the data acquisition has a smaller redundancy. Signal sparsity and sampling of irrelevance is both the premise of the CS method. Among them, the signal sparsity refers to the signal sparse representation which has great influence on the signal reconstruction in SHM. This paper discusses a method of signal sparse based on wavelet basis and analysis the signal sparse performance for SHM under different wavelet orthogonal basis. Experimental results show that the sparse decomposition performance is best when the wavelet orthogonal basis chose db5 or db6. Data analysis showed that signal sparse under db5 or db6 can ensure signal reconstruction accuracy, improve the compression rate and reduce data and transmission cost.