Modulation Recognition And Parameter Estimation Method Of PSK Signal Based On Compression Sensing

Based on the Compressed Sensing (CS) theory, the methods of automatic modulation recognition (AMR) of PSK signals, as well as the estimations of their carrier frequency and symbol rate are studied in this article. In this paper, based on the theoretical analysis, we establish the sparse model of power spectrum density of PSK signals’high order statistics. Besides, we propose the models of recovering spectrum of high order statistics of PSK signal based on CS theory, the model of recovering high order statistics of PSK signals based on low-rank matrix completion theory, and the finite-alphabet model based on CS theory. As to the models of recovering the spectrum of high order statistics of PSK signal, we propose two different schemes, according to the processing procedure and measurement matrices.The models we proposed in this paper are guaranteed by theoretical analysis, and are validated by simulations. Thanks to the existing of noise, the data matrices of PSK high order statistics generally have no obvious low rank structure. To solve the problem, we propose an efficient modified algorithm.