Research On High Resolution Azimuth Spectrum Estimation Algorithm Of AVS Based On Fourth Order Cumulant

The underwater acoustic field has the characteristics of both scalar field and vector field.The acoustic vector sensor can output the sound pressure and vibration velocity simultaneously and in common point and obtain higher processing gain.Due to the limitation of carrier volume,small platforms such as buoys cannot be installed with large aperture arrays,so a single acoustic vector sensors or small aperture acoustic vector arrays are more advantageous.Compared with the sound pressure sensor,a single acoustic vector sensor can achieve unambiguous target azimuth estimation.The invention and application of the acoustic vector sensor provides a new idea for underwater acoustic signal processing and a new approach for underwater acoustic target detection,localization and identification.Therefore,this paper studies the target azimuth spectrum estimation algorithm based on the acoustic vector sensor.The specific contents are as follows:Firstly,aiming at the problem that a single-vector sensor cannot distinguish multiple targets,a single vector azimuth spectrum estimation algorithm based on the fourth order cumulant is proposed.The paper introduces the definition and properties of the fourth order cumulant,which determines that the fourth order cumulant can effectively suppress Gaussian noise.The idea of the fourth order cumulant algorithm is applied to the single vector azimuth spectrum estimation,and the virtual aperture expansion of the single vector sensor is realized,so that it can realize the azimuth estimation of multiple targets.The results show that the fourth order cumulant processing can effectively improve the orientation estimation accuracy and multi-target discrimination ability of the single vector sensor.Then,aiming at the problem of low resolution of broadband azimuth spectrum of single vector sensor,this paper proposes a single vector broadband azimuth spectrum estimation algorithm based on sub-band peak energy detection.The traditional incoherent subspace algorithm divides the received signal into each sub-band,and then superimposes the azimuth spectrum of each sub-band to obtain a broadband azimuth spectrum,and its azimuth resolution is limited.The sub-band peak energy detection algorithm can strengthen the signal in the target azimuth and suppress the noise interference in other azimuths,thereby effectively improving the resolution of the broadband azimuth spectrum of the single vector sensor.The performance of the proposed algorithm is verified and analyzed by simulation experiments,and the results verify.Finally,the fourth order cumulant processing and sub-band peak energy detection algorithm are applied to the azimuth spectrum estimation of uniform vector linear array.Compared with scalar array,there is no "left and right blur" phenomenon,and the azimuth estimation accuracy is higher.The virtual expansion of the array based on the fourth order cumulant can effectively improve the azimuth estimation accuracy,and the sub-band peak energy detection algorithm further improves the resolution of the broadband azimuth spectrum.The performance of the proposed algorithm is verified and analyzed by simulation experiments,and the results verify the effectiveness of the proposed algorithm.