Non-cooperative Frequency Hopping Signal Processing Technology Based On Blind Compressed Sensing

In the 21 st century, the form of war has transformed gradually from mechanization to the information, and the confrontation is no longer limited to strategic equipment, accurate and efficient information processing tools become a critical factor in determining the direction of war. Non-cooperative frequency hopping signals have well-documented merits military applications due to their near-far resistance and robustness to jamming. To find effective interception, identification and recovery algorithm is one of high-profile issues in the current field of military communications. With hardware facilities sophisticated, non-cooperative frequency hopping signals gradually to develop in the direction of “High speed, Multiple frequency, Ultra wideband". The waveforms of signals become more complex, but the priori information was fewer, these pose an unprecedented challenge to modern signal processing methods.Blind compressed sensing is a sampling reconstruction theory based on the process of projection measurement, sparse coding and base transforming. On the one hand, the theory retains the characteristic of compressive sampling of traditional compressed sensing theory, extracting useful information of signals by projection measurement. And it compresses data maximally without destroying the signal structures. On the other hand, the theory fuses the process of dictionary learning, find a pair of sparse matrix and transform base with optimal product by alternately iterative, and restore the original signal structures. Blind compressed sensing theory can reconstruct signals without priori information, and provide important theoretical support for the development of non-cooperative communications.This paper studies on the technology of non-cooperative frequency hopping signal processing based on blind compressed sensing, including signal model establishing, signals reconstructing and parameters estimating. This paper introduces the development process of frequency hopping technology firstly, and summarizes the existing processing methods of frequency hopping signals. It describes the proposed process of blind compressed sensing and the corresponding theorem, summarizes three uniqueness constraints and corresponding optimal algorithms, and analyzes the characteristics of each of the three algorithms. Then introduce non-cooperative frequency hopping communication system, important parameter index and modulation scheme. According to the definition of non-cooperative frequency hopping signals, simulation results were acquired by selecting the appropriate parameter values. The paper introduces blind compressed sensing theory to reconstruct the signal structures, and analyzes experimental error according to experimental data. Finally, a parameter estimation method is the proposed based on accurate reconstruction. The paper studies the characteristics of non-cooperative frequency hopping signals in the frequency domain, and analyzes the factors influencing the quality of estimates according to simulation data.