Research On Underwater Acoustic Pulse Signal Detection Technology Based On Fractional Fourier Transform And Multichannel

Checking the presence or absence of signals is the first and most important step in signal processing.At present,most of the classical algorithms are used to detect continuous wave signals,but the detection performance of time-varying signals often cannot meet the needs of practical applications.With the development of sonar technology and underwater acoustic engineering technology,human beings continue to strengthen scientific and technical activities such as marine resource exploration,marine biological investigation,underwater acoustic target detection,and underwater acoustic communication.Various underwater sonar platforms often receive various acoustic signals,and it is of great significance to detect these acoustic signals to further obtain the information contained in the signals.This paper mainly focuses on the blind detection technology of continuous wave and frequency modulation signals when the parameters are unknown.To improve the overall performance of detecting different types of signals at a low signal-to-noise ratio,this paper studies the blind detection technology of underwater acoustic weak signals from two aspects:fractional Fourier transform and down-sampling conversion multi-channel correlation superposition processing.Common algorithms for blind detection of underwater acoustic signals with low SNR are given.Through computer simulation,the classical blind detection algorithms such as energy detection,Power Law detection,Improved Power Law detection,short-time cross-correlation detection,and entropy detection are analyzed.By comparing and testing the detection performance of each algorithm under continuous wave signal and linear frequency modulation signal,the influence of window length,window function superposition,and different quantization methods on the detection performance of each algorithm is studied.The principle of signal detection based on fractional Fourier transform is introduced,and the research status,advantages,and disadvantages of three different discrete algorithms are compared and analyzed.Aiming at the unknown frequency modulation slope,two classical algorithms for finding the optimal order are studied,and a fast order search method based on delayed autocorrelation is given.Through computer simulation,the detection performance of short-time fractional Fourier transform detection algorithm in detecting continuous wave signal,linear frequency modulation,and hyperbolic frequency modulation signals are analyzed.The application of the phase rotation spectrum correlation algorithm in single-channel signal detection is studied.Aiming at the problem that the detection performance of the classical algorithm decreases when processing Frequency Modulation signals,a down-sampling conversion multi-channel correlation superposition detection algorithm is proposed,and the detection performance of the two algorithms in detecting continuous wave signal and Frequency Modulation signals is simulated and analyzed.Experimental results show that the multi-channel correlation superposition detection algorithm based on down-sampling conversion to multi-channel can effectively improve the detection performance of Frequency Modulation signals,and the algorithm can achieve good detection results when detecting continuous wave signal,linear frequency modulation,and hyperbolic frequency modulation,and its computational complexity is much lower than that of short-time fractional Fourier transform detection.This paper analyzes the influence of underwater acoustic channel on signal detection and interception,and compares the advantages and disadvantages of current single threshold detection methods,and proposes a double threshold detection method based on the weak mean.The experimental data of Qiandao Lake are used to test the performance of the algorithm,and the detection performance of each algorithm under different signal-to-noise ratios is analyzed.Finally,the experimental results show that the detection effect of the down-sampling conversion multi-channel correlation superposition method is better than the traditional detection algorithm.