Linear ESPRIT-Like Algorithms for Fast Directions of Arrival Estimation with Real Structure

Abstract

This paper proposes two Linear ESPRIT-like algorithms for Directions of Arrival (DoA) estimation problem. The assumed constraints are the same as those imposed onto the standard ESPRIT algorithm. We introduce a new approach, for both signal subspace and eigenvalues estimation, that can efficiently replace the requirement for the classic Eigenvalue Decomposition techniques. This approach, requiring only linear operations, makes the DoAs estimator faster while maintaining comparable estimation accuracy. The proposed algorithms allow high resolution capabilities with reduced computational cost and lower processing time as compared with existing schemes. The Cramer Rao Bound on the variance of DoAs estimated by the proposed algorithms is analysed. The simulation results confirm that high resolution on DoAs estimation can be achieved by the developed methods and prove the validity of our approach.