Research On Recognition And Processing Of Composite Modulation Signals

Driven by the rapid development of satellite communication technology,the modulation techniques of measurement and control signals are constantly changing.On the basis of the traditional single modulation mode,the composite modulation signal modulated twice or more has attracted much attention of scholars.In this thesis,based on the traditional modulation signal research,for PCM/BPSK,PCM/QPSK,PCM/BPSK/PM,PCM/QPSK/PM,PCM/BPSK+BPSK/PM,PCM/BPSK+QPSK/PM and PCM/QPSK+QPSK/PM seven kinds of composite modulation signals,modulation type identification,parameter estimation and demodulation algorithm are studied.The main work of this thesis is as follows:Firstly,according to the mathematical model of the composite modulation signal,the frequency domain characteristics and statistical domain characteristics are analyzed,and the spectrum expression and fourth-order cumulant expression are derived.Based on the square spectrum characteristics and fourth-order cumulant characteristics of the signal,a direct recognition algorithm is proposed.In order to improve the recognition performance of single subcarrier signal and multiple subcarrier signals under low SNR,this thesis uses the instantaneous phase method to extract the inner modulation signal of the composite modulation signal,and extracts the number of subcarriers of the inner signal by calculating the LPC spectrum of the signal.On this basis,combined with the characteristics of signal square spectrum,an indirect recognition method for layer-bylayer recognition of composite modulation signal is proposed.Then,the carrier frequency estimation,side-carrier frequency estimation and modulation index estimation algorithms of composite modulation signals are studied.In carrier frequency estimation,according to the different frequency domain characteristics of independent carrier signal and unified carrier signal,FFT method and spectral line method are used to roughly estimate the carrier frequency value,and CZT algorithm is used to refine the coarse estimation.In order to solve the problem that the conventional frequency estimation algorithm can not estimate multiple frequency values at the same time,this thesis proposes a subcarrier frequency estimation algorithm based on MUSIC spectrum.The estimated value of subcarrier frequency is obtained by searching the frequency corresponding to the spectral peak in the MUSIC spectrum of the signal.By analyzing the discrete relationship between the modulation index and the spectral line prominence in the spectrum,this thesis uses the interpolation method to establish the continuous mapping relationship between the two,and analyzes the estimation performance of different interpolation functions.Finally,on the basis of modulation recognition and parameter estimation,the composite modulation signal is demodulated layer by layer.The outer signal demodulation adopts forward phase unwrapping,and the inner signal demodulation adopts coherent demodulation.In order to separate multiple subcarrier signals in the inner layer,this thesis adopts the idea of sliding window smoothing filtering to obtain single subcarrier signal.Under the two different identification algorithms of direct identification and indirect identification,the demodulation error rate of the signal is less than 0.1 when the SNR is 14 d B and 16 d B,respectively.