Research And Software Implementation Of Underwater Acoustic Communication Signal Reconnaissance Technology

Underwater acoustic communication signal reconnaissance technology is an important part of modern underwater acoustic communication.It plays an important role in the fields of underwater acoustic adaptive communication,underwater acoustic electronic countermeasure and civil spectrum monitoring.Its core is the blind detection technology,parameter estimation technology and modulation identification technology of underwater acoustic communication signal.Due to the complexity of underwater acoustic channel,there are still many key problems in underwater acoustic communication signal reconnaissance.This thesis studies eight kinds of underwater acoustic communication signals,including single frequency(CW),linear frequency modulation(LFM),binary phase shift keying(BPSK),orthogonal phase shift keying(QPSK),binary frequency shift keying(2FSK),quaternary frequency shift keying(4FSK),direct sequence spread spectrum-binary phase shift keying(DSSS-BPSK)and orthogonal frequency division multiplexing(OFDM).Blind detection technology,carrier frequency estimation technology and modulation recognition technology of these eight communication signals are deeply discussed.Aiming at the blind detection technology of underwater acoustic communication signal,three classical methods,spectral entropy detection,wavelet detection and short-time Fourier transform detection,are studied in this thesis,and the short-time Fourier transform detection is improved by using the sum idea.The simulation results show that the improved short-time Fourier transform detection method is superior to the three classical methods in the Qiandao Lake multipath channel,and it improves the whole signal by about 2 d B compared with the classical short-time Fourier transform detection method.Finally,based on the improvement of the short-time Fourier transform detection method,the start and end points of the signal are extracted using the double-threshold endpoint detection algorithm.Aiming at the carrier frequency estimation technology of underwater acoustic communication signals,this thesis studies the carrier frequency estimation technology based on five different methods,including Welch transform,M-power spectrum,AR(Autoregressive)model,cyclic spectrum and periodic graph.Based on the above study,aiming at the low accuracy of carrier frequency estimation of 4FSK signal in underwater acoustic multipath channel,a carrier frequency estimation method based on spectral line decision is proposed.The simulation results show that the proposed 4FSK carrier frequency estimation method based on spectral line judgment is superior to the classical carrier frequency estimation method in the Qiandao Lake multipath channel.Eight kinds of carrier frequency estimation schemes for communication signals are given.Aiming at the modulation recognition technology of underwater acoustic communication signal,this thesis proposes two sets of modulation recognition schemes of underwater acoustic communication signal,which are multi feature fusion modulation recognition scheme based on SVM(Support Vector Machine)and multi feature layered modulation recognition scheme based on deep learning joint SVM.The simulation results show that the latter has better performance,in addition to the slightly reduced recognition performance of 4FSK and DSSS signals,the modulation recognition rates of CW,LFM,BPSK,QPSK,2FSK and OFDM signals have been improved to varying degrees.Finally,the performance of the proposed underwater acoustic communication reconnaissance algorithm is verified by sea test data,and the underwater acoustic communication signal reconnaissance software is developed based on QT 5.9 and visual studio2013 development platform.