| When the wireless sensor network samples the monitoring area through limited nodes,it is usually unable to obtain the information in the area completely.Aiming at this problem,a grid data reconstruction algorithm is proposed,which uses compressed sensing technology to recover the information in the area completely.The improvement of the reconstruction algorithm is completed,and the measurement matrix based on empirical mode decomposition is designed,which effectively improves the reconstruction accuracy and reduces the running time.This paper mainly conducts in-depth research on the signal reconstruction algorithm and measurement matrix design in the compressed sensing theory.The main work includes:(1)Aiming at the signal under-estimation or over-estimation problem caused by the fixed selection step in the iterative process of the sparsity adaptive matching pursuit(SAMP)algorithm,a threshold-based sparsity adaptive matching pursuit(TSAMP)algorithm is proposed.It introduces the idea of threshold value into atomic selection,and proposes a variable step size strategy based on the change of step size,and reconstructs the signal with different step size estimation sparsity according to different conditions.(2)Aiming at the problem of too many free elements in the traditional random measurement matrix and the difficulty of realization,a measurement matrix based on empirical mode decomposition is designed.The characteristics of the signal are used to obtain the matrix components through empirical mode decomposition,and the cyclic displacement method of cyclic matrix is used for reference.A suitable measurement matrix was obtained,and its advantages compared with the traditional measurement matrix were verified by simulation,which further improved the accuracy of signal reconstruction.(3)Aiming at the limitation of obtaining complete information in the area through limited nodes in environmental monitoring,a grid data reconstruction algorithm is proposed.It combines the TSAMP algorithm and the empirical mode decomposition measurement matrix to complete the sparse sampling of the area by the limited sensor nodes,and then realizes the efficient reconstruction of the environmental information in the area to meet the requirements of environmental monitoring.This paper chooses MATLAB simulation platform to design experiments,uses classic algorithms as a reference to realize the comparison between the grid data reconstruction algorithm and the classic algorithm in the reconstruction effect of the monitoring area.The simulation results show that the improvement effect of the grid data reconstruction algorithm is better,the important characteristics of the original signal can be reconstructed more completely,and the data changes are relatively stable,which can better characterize the environmental conditions of the monitoring area. |
