The Research On Recognition Of Digital Modulation Signal Based On Cyclostationary Theory

The Significance of automatic identification system of digital communication modulation signal in a complex communication electromagnetic environment is that it can subsume an unknown modulated signal under the appropriate form of modulation types based on the rare prior knowledge. The identification and classification processing made to signals primarily provide a reliable basis for the subsequent further analysis and processing of the signal. In many applications, such as spectrum monitoring, emergency rescue, modulation recognition technology has been widely used, and the process has a very important role. In recent decades, scholars have done a lot related research in this direction and also made a number of reliable modulation recognition algorithms which are used to do recognition processing for the modulation signal. But the study finds that the efficiency of many algorithms is not good under complex communication electromagnetic environment. For this case, based on some of the latest research results, this dissertation study in a complex electromagnetic environment how to use some of the differences of signals to identify the digital modulation signals based on the characteristics of the digital communication modulation signals. The research results of this dissertation have certain role for modulation recognition theory and practice, as well as subsequent research foundation.Under this background, The work and innovation including the following:1. We research and analyze the classical statistical characteristics, based on the characteristics of the information, we divide those into two broad categories:the characteristics based on transient information and slowly varying information.2. Through the study of the power spectrum of the modulation signals that is transform by a square nonlinear system, we find that different signals would appear different discrete spectral lines. In this dissertation, wo use AR model to extract the spectral features, then use these feature to identify the modulation signals. We also act the MATLAB simulation.3. Based on the study of the cycle stationary of the modulation signals, we modeled the digital modulation signals to LPTV models. Then we estimate the model structure based on the correspondence between the LPTV models and signals. In the end, we identify the signals.