Smart Antenna System Implementation under Multipath Propagation Using JADE-MVDR and LMS Algorithms

Abstract

This paper considers the implementation of smart antenna system under multipath propagation. Here, it is considered different non-coherent signal groups each containing direct and multipath signals. The direction of arrival (DOA) of all the signals in each group is estimated using Minimum Variance Distortionless Response (MVDR) in conjunction with joint approximate diagonalization of eigenmatrices (JADE) algorithm. The generalized steering vectors are first estimated using JADE algorithm, and then the MVDR method is realized to estimate the DOA of each signal. The computation times of JADE-MVDR and JADE-MUSIC algorithms are compared for a single iteration and the results show that JADE -MUSIC has slightly lower runtime. Besides, RMSE performances are compared for different scenarios and JADE-MVDR is found to be more effective. The DOAs obtained are then processed using LMS adaptive beamforming algorithm to steer the main lobes of the radiation pattern toward the signal of interest angles and the nulls toward the signals not of interest angles. In addition, a new measure of the power level reduction under different scenarios (snapshots and array elements) is presented. The simulation results reveal that a maximum power drop of 0.4 dB is observed, and adaptive beamforming is successfully done by mitigating the effects of multipath significantly.