Fast estimation of signal subspace based on multi-stage Wiener filter

A reduced rank method for estimating signal subspace was proposed,which was based on the principle of multistage Wiener filter.Without any prior knowledge of desired signals,the signal subspace can be quickly estimated by utilizing the proposed method.Theoretical analyses and experiments show that,compared with the existing approaches to signal subspace estimation,the proposed method has not only lower computational complexity,but also better estimation performance,which means that it can greatly meet the needs for real-time processing of signals in engineering applications.