Nonlinear adaptive channel equalization using genetic algorithms

Abstract

Nonlinear adaptive channel equalization is a well-documented problem. Equalizers based on the complex decision feedback recurrent neural network (CDFRNN) have been intensively studied to address this problem. However, when trained with conventional training algorithms like the real time recurrent learning (RTRL) technique, the equalizer suffers from low convergence speed, requiring very long training sequence to achieve proper performance. In this work, we propose a new approach to equalize nonlinear channels using genetic algorithms. The proposed Volterra decision feedback genetic algorithm (VDFGA) uses a genetic optimization strategy to estimate Volterra kernels in order to model the inverse of the channel response. Simulation results show very high convergence speed, which allowed to achieve interesting bit error rate (BER) using relatively short training symbols, when considering only 8-bits long coded weights.

Bayesian Volterra system identification using reversible jump MCMC algorithm

“…These methods generally perform Volterra system coefficient estimations based on nonlinear least mean squares (NLMS), least mean pth power (LMP), nonlinear recursive least squares (NRLS) and extended Kalman filters [7,[15][16][17][18][19][20][21] . Furthermore, genetic algorithms [22,23] , QR decomposition [24] , neuro-fuzzy [25] and neural network [26] architectures have also been used in VSI studies. For all these studies, nonlinearity degree of Volterra model is assumed to be known.…”

Swarm Intelligence Based MMSE Frequency Domain Equalization for MIMO Systems

“…The local minima problem in gradient-based optimization restricts its performance [2], mainly in non-linear channel conditions. The proved efficiency of derivative-free optimization algorithms like Genetic Algorithm (GA) [5], Artificial Bee Colony Optimization (ABC) [7], Simulated Annealing (SA) [6], and PSO [2] [8] [9] triggered the search for intelligence-based swarm optimization in channel equalization over the last decades. The common process steps such as selection, mutation, crossover, reproduction, group best experience, and personal best experience lead to the popularity and performance of these algorithms at the higher level of wireless communication applications.…”