Sparse signal reconstruction method based on IPNN and compressed sensing

Abstract

The invention belongs to the technical field of signal reconstruction, and discloses a sparse signal reconstruction method based on IPNN and compressed sensing. Through the thought of model conversion, the problem of solving unconstrained Lq minimization is converted into a problem of solving variational inequality on a convex set, and then on the basis of existing research, an inertial projection neural network is successfully used for solving the optimization problem and strictly proved; and according to the designed signal reconstruction algorithm, a series of abundant numerical simulation experiments are carried out, the experiment result shows that the algorithm is feasible in sparse signal reconstruction, and compared with an existing L1 and Lq minimization solving algorithm, the algorithm is better in reconstruction performance under noise interference.

Description

technical field [0001] The invention belongs to the technical field of signal reconstruction, and in particular relates to a sparse signal reconstruction method based on IPNN and compressed sensing. Background technique [0002] At present, the 21st century is an era of rapid development of information science theory and technology. With the explosive growth of information volume, the continuous expansion of information demand and the diversification of information types, the traditional methods of signal sampling, transmission and storage are facing great challenges. At the same time, with the continuous improvement of computer performance and the continuous deepening of signal transmission theory and technology, the field of information science is stirring waves of innovation one after another! [0003] The simulation (virtualization) of real signals and the digitization of signal processing determine that the signal sampling process is the only way to obtain digital sign...

AI Technical Summary

Problems solved by technology

However, the existing traditional sparse signal reconstruction algorithms are mostly supported by classical optimization theory and methods, and its optimization and processing performance is difficult to improve, and its reconstruction performance lacks robustness in the case of noise interference.

[0010] To sum up, the problem existing in the existing technology is: the reconstruction performance of the existing traditional sparse signal reconstruction algorithm in the case of noise interference needs to be improved urgently

[0011] The difficulty in solving the above problems and defects is: how to integrate the neural network method with powerful optimization and processing performance into the design of the sparse signal reconstruction algorithm, especially transform the solution problem into a form that is easy to optimize for the neural network, and in addition , how to improve the reconstruction performance in noisy situations is also one of the difficulties

Images