A method for spatial signature estimation with uncalibrated uniform linear array

In this paper, a multidimensional search procedure is presented for estimating spatial signature with uniform linear array in the presence of unknown gain and phase errors. The proposed method fully exploits multiple invariances in the interior of the sensor array and it is desirable that the quality of estimate improves as more and more array invariances are used. A Gauss-Newton iterative algorithm is applied for solving the multidimensional nonlinear problem. Simulation results indicate that a solution close to the minimum is obtained after only two or three iterations. Furthermore, the introduction of a weighting matrix, similar to weighted subspace fitting, can tone up the performance of our method.