Performance Analysis of 2D-ESPRIT Algorithm for URA in Estimation of 2D-DOA in Massive-MIMO Systems

Abstract

Purpose: The performance of smart antenna is affected by antenna array configuration and the direction of arrival estimation (DOA) algorithm used .To have a suitable DOA for an uniform rectangular array (URA),The performance of two dimensional estimation of signal parameters via rotational invariance techniques (2D-SPRIT) that are applied for two different planar dispositions of antenna for URA in Massive Multiple Inputs Multiple Output (Massive MIMO) systems when   subjected to the same simulation conditions is comparatively analyzed. The performance of the proposed 2D-ESPRIT algorithm with URA in vertical plane (the case of full-dimensional MIMO) has been compared with that of the existing 2D-ESPRIT algorithm that is applied with URA in horizontal plane. Methodology: Implementation and simulation of analyzed 2D-ESPRIT algorithms was done in MATLAB and compared considering computational complexity efficiency, DOA estimation precision, estimation failure and variation of Root mean square Error (RMSE) with Signal to Noise Ratio (SNR) and angular spread. MATLAB simulation results got from similar input parameters in a multipath environment are compared to derive a conclusion. Findings: Simulation results show that the proposed 2D-ESPRIT algorithm applied in vertical plane outperforms the existing 2D-ESPRIT algorithm applied in horizontal plane for all considered performance criteria that include estimation precision, running time and Root mean square error (RMSE). Recommendation: Other researchers interested in working on 2D-ESPRIT algorithms are recommended to improve the work reported in this paper by further reducing computational complexity observed in 2D-ESPRIT algorithm with URA in vertical plane at large antenna array size to reflect the standard of full-dimension MIMO.