The Research Of 3-D Forward Looking Sonar On High Resolution DOA Estimation

Three-dimensional forward-looking sonar is a necessary guarantee for the safety of the Autonomous Underwater Vehicles (AUV) and Submarines. A 3D forward-looking sonar is more versatile than conventional 1D and 2D ones, which can provide the direction, distance and depth information of the obstacles ahead of the platform. The paper mainly discusses and researches several methods on the direction-of-arrival (DOA) estimation of the targets' echo.FFT beamforming method and the shifted sideband beamforming method are reviewed. The Bartlett beamformer and MUSIC method based on fourth-order cumulants to the vector-sensor array are presented for the first time. The main interest in using fourth-order cumulants instead of second-order ones in vector-sensor array processing application is the aperture extension property of higher-order cumulants, which makes it possible to increase both the resolution and the number of sources to be processed from a given array. This method makes it possible to miniaturize the shape of the array. Through computer simulation and trial data, this method is feasible.Spaces smooth methods are used to reduce the correlation between the echo signals to ensure the accuracy of the estimation of targets number and higher resolution of the DOA results. Considering both complexity and calculation precision of most high resolution methods, ESPRIT and CAATI algorithm are mainly introduced and discussed. According to the computer simulation, ESPRIT algorithm has both good stability and precision, and CAATI algorithm can give both the direction and the amplitude of the echo, but it is needs high signal-to-noise ratio. And the two algorithms are verified to be feasible by dealing with the experimental data from lake at the last part of this paper.