The Processing Of Simple Shape Underwater Targets Echo Based On The Wigner Higher-order Spectrum

For active sonar targets detection and identification as a hot spot of the underwater acoustic field, how to detect target echo from the received signal, and extract feature information from echo is its foundation and also its key technique. Based on the distribution characteristics of the target echo and the background noise in the time-frequency domain and the time-frequency difference between target echo highlights, we prefer to use time-frequency methods to deal with these non-stationary signals. With the increase of the interest for the higher order spectrum,higher order spectral analysis has become a important tool to analysis non-gaussian process as well as extract more signal spectrum information. However it is not suitable for non-stationary process. Thus in this paper, it proposes to use the time-varying higher-order spectrum, i.e. the Wigner Higher-Order spectrum to achieve target echo detection and bright spot extraction.This paper deduces the Wigner bispectrum and Wigner trispectrum of the CW signal based on the definition of the Wigner higher-order spectrum, and studies the detection capabilities of Wigner higher-order spectrum for the CW signals and LFM signal in the background of gaussian white noise . It also studies the cross terms of the Wigner bispectrum and Wigner trispectrum of multi-component CW signals and LFM signal, and gives the characteristics of cross terms for different signal. For the shortcomings of using kernel function to depress cross terms, according to the basic algorithm of morphological processing,a method combining the features of cross-term of Wigner bispectrum and Wigner trispectrum characteristics of the multi-component signal with the morphological processing ,is proposed in this paper to deal with the cross-term. In addition, according to the distribution characteristic of the noise and signal in the Wigner higher order spectral domain, it is pointed out that the method can also be used to noise suppression processing.Based on target echo highlights model, this paper presents the number of the target echo highlights and their time delay in different incident angles of a typical underwater target as well as analyses the higher-order time-frequency characteristics of the target echo. Based on the cross processing method of morphology, it uses the Wigner high-order spectrum to processing the target echo highlights. The simulation analysis gives the scope of application of the method.Finally, according to the results of the experiment data processing, the accuracy of target echo highlights model was verified, the method of Wigner higher-order spectra combined with morphology in geometric point extraction is proved feasible. Also this method is an effective tool to detect the target echo.