Fast joint DOA and inter-subarray errors estimation in partly calibrated arrays

In this paper, we utilize the original noise subspace and its conjugate to reconstruct a new estimator for the fast joint estimation of DOA and inter-subarray errors in the partly calibrated arrays (PCA), namely fast-rank-reduce (F-RARE). Based on the new estimator, the collection of true DOA estimation and its virtual mirror can be obtained by searching over only half of the total angular field-of-view. Then, the true DOA estimation can be selected by using the conventional beamforming (CBF) technique. Additionally, the inter-subarray errors can also be computed by a finite number of subspace decompositions. Numerical simulation illustrates that compared with the conventional RARE algorithm, the new method can not only reduce the computational burden by a half but also provide a similar estimation accuracy.