Shallow Sea Acoustic Signal Blind Processing Based On LFM

The LFM signal has good autocorrelation characteristics and wide Doppler tolerance.It is often used as a synchronous beacon for underwater acoustic communication.By analyzing and processing the underwater acoustic LFM signal,it helps to find the signal and obtain relevant water.The information of the parameters of the acoustic channel,and the channel parameter information is used in the signal demodulation process to improve the blind demodulation performance of the underwater acoustic communication signal.Under the white Gaussian noise condition,underwater acoustic LFM signal identification,parameter estimation and blind demodulation have been a lot of achievements.However,underwater acoustic channels are complex and varied,especially in shallow seawater acoustic channels,which are also accompanied by a large amount of impulse noise,making existing The performance of the underwater acoustic LFM signal analysis and processing in the Gaussian white noise environment is degraded or fails.In this paper,for the shallow seawater acoustic non-cooperative communication signal with impulse noise,the recognition and parameter estimation of the LFM signal in the complex underwater acoustic channel under the background of impulse noise are studied.Then the channel parameter estimation based on the LFM signal in this environment is studied,and then the underwater acoustic MPSK is designed.The signal blind demodulation scheme gives each link algorithm and finally completes the effective blind demodulation of shallow seawater MPSK communication signals.The main work and research results of the thesis are summarized as follows:Firstly,for the problem of LFM identification under the environment of low SNR impulse noise,a method for identifying shallow seawater acoustic LFM signals based on nonlinear mapping discrete fractional Fourier transform is presented.The algorithm first performs nonlinear transformation on the received signal,suppresses the impulse noise,then performs discrete fractional Fourier transform,designs a normalized variance ratio as the signal recognition feature,and finally completes the recognition of the LFM signal through a support vector machine.Compared with the existing algorithms,this algorithm avoids the difficulty of selecting the threshold when judging by the threshold,and improves the recognition rate when the signal to noise ratio is low.The effectiveness of the algorithm is verified by simulation experiments and field experiments.Secondly,a parameter estimation method of LFM signal based on two-dimensional particle swarm optimization is used to estimate the parameters of LFM signal in the low-SNR environment.Based on the rough estimation of the LFM signal parameters based on the discrete fractional Fourier transform,this algorithm further designs a two-dimensional particle swarm optimization method to make a fine estimation.Compared with existing algorithms,this algorithm has higher accuracy under similar computational conditions.The effectiveness of the algorithm is verified by simulation experiments and field experiments.Finally,aiming at the problem of blind demodulation of non-cooperative MPSK signals in shallow sea under impulse noise environment,a blind demodulation method based on LFM signal estimation and processing is given,and the algorithm of each link is given.The algorithm firstly performs the nonlinear transformation based on the LFM signal strength to suppress the impulse noise,which reduces the influence on the signal while suppressing the impulse noise signal better.Then,the LFM signal is subjected to the fractional Fourier transform and analyzed,the tap coefficients of the fractional blind equalization are initialized by the obtained channel parameter information,and the signal is finally adjusted by the M-th power transformation and the phase-locked loop to adjust the frequency offset.Demodulation.The effectiveness of the algorithm is verified by simulation experiments.