Reflection equivariant Crayfish quantum neural network based channel estimation in massive multiple input multiple output OFDM

Abstract

This study presents Reflection equivariant Crayfish Quantum Neural Networks (RCQNNs), reflection equivariant, optimized through the Crayfish Optimization Algorithm, toward efficient channel estimation, in a massive multiple-input, multiple-output orthogonal frequency-division multiplexed (MIMO-OFDM) system spectrum. Massive MIMO and OFDM have become two very important technologies in modern wireless communications, improving the efficiency of spectral and data rates. High-dimensionality of wireless channels has also made accurate channel estimation a significant challenge in such systems. In this respect, the research projects Reflection Equivariant Quantum Neural Networks (REQNNs) for using quantum neural networks (QNNs) to reflect equivariance in the network, capturing how symmetries appear inside a wireless channel. The REQNN framework guarantees that reflection symmetry holds in the model, which should be seen to generalize better and show higher prediction competency of channel states. The system further optimizes the Crayfish Optimization Algorithm to mimic crayfish behavior in nature as an input for fast fine-tuning of network parameters. Through it, the solutions are converged quickly, thus boosting network operation efficiency. Simulation results reveal that the proposed REQNN outperforms conventional models like CNN, MMV-LAMP, and FPN in MSE, PSNR, and computational efficiency. The proposed REQNN also achieves up to an 80% MSE reduction compared to existing methods and delivers up to 20 dB of PSNR increment, thus better reflecting its superior accuracy in channel estimation. Moreover, cryptography's time consumption during encryption and decryption is reduced to 1.42 and 1.34 s, respectively, which further streamlines real-time applicability in real-world situations for the MIMO OFDM channels.