Performance of adaptive beamforming using the split RLS algorithm

Abstract

In this paper a novel adaptive beamforming scheme is described. This scheme is implemented basically based on the GSC (generalized sidelobe canceller) algorithm which is one of the famous adaptive interference suppression schemes. The implementation of GSC requires complex computations due to the adaptive filtering. Therefore, we propose an efficient GSC employing the split RLS algorithm. However, this method leads to an estimation error caused by the correlation between the split vectors. In order to reduce the estimation error the proposed GSC is combined with a Walsh blocking matrix instead of a bi-diagonal one. In conclusion, the SINR of the proposed method is as close as possible to the SINR obtained by full tap solution as its complexity decreases. When the system complexity decreases nearly half the SINR of the proposed method becomes 1 dB worse than that of the full tap solution.