Abstract:

A robust direction-of-arrival (DOA) estimation algorithm against arbitrary impulse noise of unknown statistics is proposed. The algorithm extends the concept of spatial sign and rank in nonparametric statistics so as to construct an array covariance matrix by using the normalized spatial sign function， then a subspace method is applied to direction finding. It is feasible to estimate DOAs according to the normalized spatial sign covariance matrix. On the basis of the theoretical proofs the algorithm can be generalized to any subspace methods. Computer simulation results show that the algorithm is robust in the arbitrary impulse noise environment and is not sensitive to the array gain and phase errors.