Chicken particle swarm optimization algorithm for MMSE equalization in MIMO systems

The high density of wireless devices operating close to populated areas can result in significant Inter-Symbol Interference (ISI). Further, Co-channel interference from neighboring networks and devices can disrupt communication and might decrease signal quality. Thus, it necessitates the construction of a channel equalizer to minimize the effects of inter-symbol interference (ISI) in wireless communications. The problem alleviates in multipath wireless channels, where a high data rate is required and a low channel error rate. The conventional gradient adaptive algorithm used for training the equalizer fails to address the problem as it gets caught in local optimal solutions. Recent systems have investigated the use of swarm intelligence algorithms for training purposes, and they outperform the gradient algorithm-based equalizers for severe non-linear distortion. Motivated by the above discussion, in this paper, we propose a Chicken Particle Swarm Optimization (CPSO) algorithm for training channel equalizers. The algorithm is applied for adaptive channel equalization in a 2 × 2 Multiple-input-Multiple-output (MIMO) flat fading Rayleigh multipath channel, which makes it effective in escaping the local minima problem of training the equalizer coefficients. The optimal values of the CPSO parameters are determined by performing the sensitivity analysis. Simulations prove that the proposed CPSO algorithm performs better in terms of Minimum Mean Square Error (MMSE), convergence, and Bit Error Rate (BER) compared to other metaheuristic algorithms such as Particle swarm Optimization (PSO), Cat Swarm Optimization (CSO) and Chicken Swarm Optimization (CHSO). The equalization performance of the CPSO-based training scheme was demonstrated using Linear Transversal Equalizer (LTE) and Decision Feedback Equalizer (DFE). The CPSO-DFE provides the optimum MMSE of -90 dB and BER crosses 10–4 for 30 dB Signal to Noise Ratio (SNR) whereas for PSO it crosses 10–3.