Modulation Classification And Parameter Estimation Of Communication Signals

The modulation classification of common communication signals have a significant role both in military communications,and in civilian communications.In military communications,modulation recognition can efficiently identify military communication signals,,destroy enemy signals and so on.In civilian communications,it helps to the investigation of interference signals.In the field of communications,mainly are non-cooperative communication,the information of the signal transmission both sides have is asymmetric,in order to effectively demodulate the original signal,the receiver need to know the exact parameters of the signal,which is helpful to the subsequent processing.So the parameter estimation of signals is necessary as well.The project relies on the cooperation of scientific and technological projects,with the aim of designing a complete set of signal modulation recognition and parameter estimation system for electromagnetic compatibility testing environment,including background data processing and the presentation of reception data.As for the recognition algorithm used for modulation of communication signals,this paper combines the algorithm of time domain's feature extraction and that of frequency domain.It uses the power spectrum of frequency-domain signal as the characteristic,and a hilbert transform of time-domain to construct the analytic function,recognitionrate is higher than 90%in the case of 10dB SNR.In the field of parameter estimation,this paper present a complete set of SNR estimation algorithm based on singular value decomposition and second-order four-order moments estimation,a method of signal carrier frequency estimation based on cyclic spectral estimation,a method of signal symbol rate estimation based on Haar wavelet transform algorithm,and do simulation with the real signal,the errors are within the allowable range.In addition,a front display interface is designed to show the experiment results,at the same time,a traceable signal and image processing software is wrote to identify excessive frequency.