Performance Analysis of 5 G Wireless Hybrid Precoding Using Evolutionary Algorithms

Abstract

Emerging 5G communication solutions utilize the millimeter wave (mmWave) band to alleviate the spectrum deficit. In the mmWave range, Multiple Input Multiple Output (MIMO) technologies support a large number of simultaneous users. In mmWave MIMO wireless systems, hybrid analog/digital precoding topologies provide a reduced complexity substitute for digital precoding. Bit Error Rate (BER) and Spectral efficiency performances can be improved by hybrid Minimum Mean Square Error (MMSE) precoding, but the computation involves matrix inversion process. The number of antennas at the broadcasting and receiving ends is quite large for mm-wave MIMO systems, thus computing the inverse of a matrix of such high dimension may not be practically feasible. Due to the need for matrix inversion and known candidate matrices, the classic Orthogonal Matching Pursuit (OMP) approach will be more complicated. The novelty of research presented in this manuscript is to create a hybrid precoder for mmWave communication systems using metaheuristic algorithms that do not require matrix inversion processing. The metaheuristic approach has not employed much in the formulation of a precoder in wireless systems. Five distinct evolutionary algorithms, such as Harris–Hawks Optimization (HHO), Runge–Kutta Optimization (RUN), Slime Mould Algorithm (SMA), Hunger Game Search (HGS) Algorithm and Aquila Optimizer (AO) are considered to design optimal hybrid precoder for downlink transmission and their performances are tested under similar practical conditions. According to simulation studies, the RUN-based precoder performs better than the conventional algorithms and other nature-inspired algorithms based precoding in terms of spectral efficiency and BER.