A Theoretical Performance Study of the ESPRIT Algorithm for the 7-element ESPAR Antenna

摘要

Recently, the Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT) algorithm for the Electronically Steerable Parasitic Array Radiator (ESPAR) antenna has been adapted for the direction of arrival (DoA) estimation of coherent signals by adding forward-backward averaging (FBA) pre-processing (FBA-ESPRIT-ESPAR). However, the performance of this algoritm has never been studied for comparison to theoretical bounds. Therefore, in this paper, the performance of FBA-ESPRIT-ESPAR is compared to those of the MUltiple Signal Classification (MUSIC) algorithm for the ESPAR antenna and the Cramer-Rao lower Bound (CRB). Simulations were performed to evaluate the performance of FBA-ESPRIT-ESPAR in terms of the number of snapshots, the correlation coefficient p, the signal-to-noise ratio (SNR), and the angular distance between the signals. Results show that, compared to the MUSIC for ESPAR, the FBA-ESPRIT-ESPAR exhibits lower performance. However, the FBA-ESPRIT-ESPAR performance is close enough to that of the CRB to consider it sufficiently efficient.