A multistage representation of the Wiener filter based on orthogonal projections
IEEE Transactions on Information Theory
Hi-index | 0.00 |
Subspace-based methods for estimating the direction of arrival are very important due to their high-resolution abilities. In this paper, a fast and efficient method based on the multi-stage wiener filter technique to estimate subspace is proposed, which doesn't require the training or reference signal. Firstly, the fact that the match filters of multi-stage wiener decomposition can span the signal subspace is simply proved. Then, a more efficient subspace approximation method is obtained. Compared with the classical eigen-vector decomposition (EVD) based method, the proposed method has less computation complexity with the approximation performance. The simulation results demonstrate its effectiveness and robustness.