Research On Iterative Reconstruction Algorithm Of Compressive Sensing

With the advancement of science and technology,a large number of sensors are put into use,and the sampling method of these equipment are designed based on Nyquist sampling theorem,although the signal can be achieved accurate reconstruction,it also brings the transmission and storage of massive data,and this sampling method depends on the bandwidth of the signal.Compressive sensing breaks out the limitation of Nyquist sampling theorem,it does not depend on the bandwidth of the signal,but on the sparsity of the signal.The reconstruction algorithm plays an important role in compressive sensing.In this paper,we focus on optimizing the compressive sensing reconstruction algorithm and improve the reconstruction performance of the algorithm.An iterative reconstruction algorithm is proposed to minimize the smoothed L0 norm and an adaptive threshold iterative reconstruction algorithm of distributed compressive sensing.The main work of this paper is as follows:An iterative reconstruction algorithm for minimizing the smoothed L0 norm is proposed.An iterative reconstruction algorithm that minimizes the smoothed L0 norm is a low complexity and high accuracy method.First,in the smoothed L0 norm,this paper takes two functions to approximate the L0 norm and solve the smoothed L0 norm problem.The gradient descent method is used to get the iterative formula.In each iteration,the iterative formula is calculated to obtain the iterative result,and the corresponding support set is obtained.Then,by using the support set to correct the iteration results,it makes the residual smaller.When the iteration stopping criteria is reached,the iteration process is stopped.Compared with the contrast algorithms,this method can achieve higher reconstruction precision,which with a low complexity method.In the case of low sampling rate for the two-dimensional Lena image reconstruction,the PSNR of the reconstructed signal is obviously higher than that of other contrast algorithms.This paper proposes an adaptive threshold iterative reconstruction algorithm of distributed compressive sensing.It is a decentralized parallel algorithm based on the distributed compressive sensing model in the distributed network.Assuming that the nodes in the network have their own computing power,each node sends the support set of its own reconstruction to the surrounding node.The node receives the supportset of the surrounding nodes,and integrates the support set.Finally,the node feeds back to the surrounding nodes,the correct support set after a number of information exchange can be obtained.The method not only effectively reduces the amount of data in the network,but also makes the reconstruction faster.Experiments show that the algorithm is applied to the wireless network.In the case of noisy,it can reconstruct the original signal perfectly.