Digital Channel Equalizer Using Functional Link Artificial Neural Network Trained with Quantum Aquila Optimizer

In digital communication, channel equalization plays an important role in mitigating the effects of inter-symbol interference, non-linearity and noise. In case of wireless channels, it also reduces the co-channel and adjacent channel interference. The channel equalizer is placed at the receiver which inherently perform an inverse modeling operation. In this manuscript, equalizers are proposed based on Functional Link Artificial Neural Network (FLANN) for nonlinear communication channel. Three types of FLANN architecture are explored based on: Trigonometric, Chebyshev and Legendre polynomial-based expansions. The weight of these FLANN architectures are trained by a quantum Aquila optimization algorithm (QAOA). In this manuscript the quantum entanglement principle is embodied to improve the performance of original Aquila optimizer. The Aquila optimizer is reported in 2021 by Abualigah et. al. is a popular algorithm and based on the inherent behavior of Aquila to catch the pray. Simulation studies are reported for two nonlinear finite impulse response (FIR) channels performance under noisy environment. The performance is reported in the form of MSE and normalized MSE (dB) value, run time required during training; final bit error rate (BER) value obtained and BER plot during testing. Simulation results reveal superior performance of Chebyshev FLANN architecture with QAOA learning, compared to the other FLANN models as well as a FIR filter-based equalizer trained with original Aquila optimizer, Grey Wolf optimizer, particle swarm optimizer and least mean square algorithm.