Underwater DOA Estimation With Variance Acoustic Velocity

Direction of arrival(DOA) estimation is a significant research area in array signal processing, which is widely used in communication, radar, sonar and navigation. In recent years, the researchers proposed many algorithms and the DOA performance is getting better and better. The DOA algorithms take advantage of the phase difference of the received signals to estimate the directions, which is closely related with the signal velocity. If the algorithm used the inaccurate velocity, it would cause errors. To overcome this problem, we focused on the DOA algorithm of velocity independent. Specific contents are as follows:1. We present the architecture structure of the DOA estimation techniques and the mathematical model of signals and sensor arrays, then introduce two classic methods:multiple signal classification(MUSIC) algorithm and estimating signal parameters via rotational invariance techniques(ESPRIT), and analyzed the effect of the sound velocity on them.2. To remove the influence of the sound velocity, two improved methods are proposed,which called VI-MUSIC(Velocity Independent MUSIC) and VI-ESPRIT(Velocity Independent ESPRIT). They both use an orthogonal array. VI-MUSIC removes the speed influence by scanning the velocity and VI-ESPRIT removes it by solving an equation. They can perform better than the MUSIC and ESPRIT when use the inaccurate velocity.3. The problems of underwater DOA estimation caused by the complexity of ocean environment are analyzed. To solve them, we combined the VI-MUSIC and VI-ESPRIT with forward and backward spatial smoothing techniques. Experimental results show the good effectiveness of them.