Modulation Classification Of Communication Signals Based On The Cyclic Spectral Correlation Function

The purpose of communication modulation signals classification is identify the modulation style of the received signals and estimate the correlative parameters on condition that there are many signals and noises, and provides the foundation for further analyzing and processing signals. This research is very useful in acknowledging the signals, distinguishing the right signals from the disturbed signals and the communication rivalry. A lot of research work has been carried out in the past few decades. But the recognition rate is very low in condition of low signal to noise ratio. According to some new thoughts home and abroad, this paper puts forward some new thoughts to recongnize the communication modulation signals.In this paper the modulation signal is modeled as cyclo stationary and analised with the cyclic spectral correlation function which represents the characteristic of the cyclo stationary signals. On account of the cyclic spectral correlation function of the modulated signals is not sensitive to noise and interference, research on the method of communication modulation signals classification in the case of low signal to noise ratio has been made.Firstly this article described the concepts which are related to the cyclic spectral correlation function, studied the characteristics of modulation signals in time domain, frequency domain, as well as the cyclic spectral frequency domain. Secondly this article improved the algorithm of the original parameter estimation about the signal of BPSK using the cyclic spectral correlation function, and gave the carrier frequency estimation algorithm about the signals of AM, DSB, 2ASK, 4ASK, 2FSK. Lastly this article used the cyclic spectral correlation theory and the parameters of statistical theory to identify the analog signals, gave the characteristic parameters and put forward the analog modulation signals classification algorithm. Used the subtractive clustering theory, gave some new features parameters by analyzing the cyclic spectral features of the digital modulation signals when do the digital modulation signals within-class classification,and enhanced the recognition rate of the modulation signals in condition of low signal to noise ratio. It combined the algorithms of with-in class and intra-class recognition about digital signals, which made the algorithm of digital modulation signals classification, and experiment recognition rate of the classification algorithm at low SNR.