Optimum DOA algorithms for narrowband and wideband applications
Haoji Bao · 2005
Direction-of-arrival (DOA) estimation is a major research topic in the field of array signal processing.In recent developments of DOA estimation, focus has been put on tailoring of existing algorithms for implementation purpose.However, there are still issues unsolved regarding the basic DOA theories, one of which is concerning phase determination/selection.Moreover, most of the existing DOA estimators were originally designed only for narrowband applications.Given the increasing need of solutions in wideband scenario, wideband DOA algorithms become important and necessary.This project targets two critical issues, 1)to study existing algorithms and to optimize them by mitigation/elimination of performance degradation caused by phase uncertainties; 2)to extend narrowband DOA algorithm to wideband application and provide wideband solutions.In this thesis, the nature of phase uncertainties in Minimum-norm (MN) and Unity-response (UR) algorithms is studied and ways of optimization are proposed accordingly.An averaged method is designed for MN algorithm by exploiting its characteristics in the concept of polynomial rooting.And an iterative way of phase optimization is proposed for UR algorithm using a constant modulus approach.Subsequently, two different methods to extend narrowband UR estimator to wideband applications are discussed.We hope that, through the thesis, we are able to make the theory of DOA estimation more complete.