Direction-of-Arrival Estimation Based on Sparse Representation of Fourth-Order Cumulants

The actual underwater environmental noise is often spatial colored, which results in severe degradation of the performance of the underwater direction of arrival (DOA) estimation method based on the assumption of white noise. In the presence of Gaussian colored noise, a high-resolution DOA estimation method using a fourth-order cumulant for quickly eliminating redundancy is adopted in this paper. Firstly, a selection matrix is constructed, and the redundant data in the fourth-order cumulants are reduced in the way of descending order. Secondly, the fourth-order cumulants matrix is transformed into a vectorized form, and the selection matrix is further constructed to eliminate redundant data in the vectorization process, and a single observation vector model with better performance is obtained. Finally, the sparse representation method is used for DOA estimation. The simulation results demonstrate that compared with the traditional fourth-order cumulant methods, this method has a stronger ability to suppress colored noise, and can provide higher resolution and higher estimation accuracy under the conditions of few snapshots and low signal-to-noise ratio. The experiment verifies that this method can be applied to DOA estimation of underwater acoustic array signals.