Efficient Channel Estimation of Massive MIMO Systems using Artificial Bee Colony Algorithm

Abstract

Massive Multiple Input Multiple Output (m-MIMO) employs a multitude of antennas at the base station to provide spectral and energy efficiency to various users simultaneously. In order to reap the full benefits of this system, the exact channel state information (CSI) is very crucial and the channel estimation is really challenging in m-MIMO scenario. In order to find a low complexity channel estimator for m-MIMO systems, we formulate the estimation as an optimization problem and to estimate the channel accurately, we propose bio-inspired Artificial Bee Colony (ABC) algorithm based method in this paper. Our study demonstrates that ABC based estimator is capable of achieving the mean square error (MSE) performance of optimal minimum mean squared error (MMSE) estimator without any knowledge of channel statistics or large scale fading coefficients. In our simulation, we also compare the performance of ABC estimator with that of Genetic algorithm (GA) based estimator. The results show that ABC estimator provides better error performance, faster convergence and less computational complexity than that of GA.