A Review of Using Artificial Intelligence to Improve Signal Reception in 5G and Future Mobile Networks

Abstract

Wireless communication systems are constantly evolving from 5th generation (5G) networks to beyond-5G (B5G) and 6th generation (6G) networks, presenting unprecedented requirements for highly reliable and efficient reception of signals. Advanced applications like autonomous vehicles, industrial automation, remote healthcare, extended reality and integrated sensing demand ultra-reliable communications with low latency (URLLC), massively connected devices and quality of service. The multipath fading and co-channel interference, dynamic user mobility, blockage and penetration losses of millimeter-wave and terahertz frequencies, and rapidly changing propagation environments, however, severely affect the reception of signals in modern wireless systems. These impairments affect signal to noise ratio (SNR), signal to interference-plus-noise ratio (SINR), bit error rate (BER), throughput, and overall communication reliability. Artificial Intelligence (AI) has become a game-changer that can tackle such challenges with data-driven and adaptive optimization. In this review, the authors comprehensively examine the roles that Machine Learning (ML), Deep Learning (DL), Reinforcement Learning (RL), Federated Learning (FL), and Explainable Artificial Intelligence (XAI) play in improving the reception of signals in 5G and future mobile networks. It is used in key applications such as channel estimation, equalization, beamforming, blockage prediction, interference mitigation, predictive handover, power control, and intelligent resource allocation. The literature provides ample evidence of the significant improvements of AI-based techniques over traditional methods in terms of SINR, BER, throughput and reliability in highly dynamic environments.