An efficient interference steering matrix estimation scheme for weight vector correction

Abstract

To make the sensor array beamformer insensitive to steering vector error, an efficient interference steering matrix (ISM) estimation scheme for weight vector correction is devised in this paper. The basic idea of this approach is converting the training data into interference-free condition via ISM estimation, thereby correcting the weight vector by the eigen-projection technique. In particular, the ISM is gained with a multi-direction desired signal (DS) blocking and matrix eigen-transition means. Subsequently, utilizing the orthogonal projector of the ISM to process the sample data, the quasi DS covariance matrix is attained indeed, which follows the extraction of the eigen-vectors having stronger correlation to the nominal weight vector. Finally, the weight vector is corrected by projecting the nominal weight vector onto the subspace spanned by the eigen-vectors above. Typical experiments verified that the offered beamforming scheme with reduced complexity, is superior to the existing methods in the situation of various steering vector errors.