Near-field source localization with partly sensor gain and phase uncertainties

Abstract

In this paper, we consider the source localization for the multiple near-field narrowband signals impinging on a uniform linear array (ULA) with partly gain and phase uncertainties. By dividing the ULA into two overlapped symmetric subarrays and assuming that they have partly unknown gain and phase responses but correspondingly identical, a new modified generalized ESPRIT based method is presented, in which the direction of arrival (DOA) and range are estimated separately. Moreover, we prove that the proposed method is equivalent to the spectral Fresnel-region rank reduction (FR-RARE) [15] method, but more computationally efficient than it. The simulations demonstrate the effectiveness of the proposed method.