Direction finding with superresolution using root implementation of eigenstructure techniques and joint estimation strategy

In this paper the problem of direction of arrival estimation when using the uniform linear array is considered. In the wireless mobile communication such problem is individual or is of interest as a part of more general problem of estimation of the mobile channel directional structure. The root modification of recently proposed joint estimation strategy (JES) is presented. The essence of JES is to use the several eigenstructure-based estimators (estimator hank) for a given sample covariance or data matrix. Appropriately combining the results of these underlying estimators, JES improves the threshold performance. The modified Root-MUSIC algorithm and modified generalized Root-Min-Norm algorithms are used as underlying estimators. The proposed root implementation of JES provides computational saving and better threshold performance than spectral JES.