Artificial Intelligence-Based Radio Access Network Optimization in 5G

Abstract

Unprecedented connection, speed, and capacity are on the horizon with the arrival of 5G networks, a major step forward in
the telecom industry. Nevertheless, 5G networks’ intricacy presents formidable obstacles that call for cutting-edge optimization
solutions. In this context, AI has shown to be a powerful tool, providing new ways to improve the performance of Radio
Access Networks (RANs). Using the following topics as a framework, this paper reviews the literature on AI-based RAN optimization:
cell placement and optimization, interference mitigation and management, resource allocation and scheduling, and
traffic prediction and management. Massive multiple-input multi-output antennas, ultra-dense networks, and millimeter-wave
communications are the three main enabling technologies for the development and implementation of 5G systems. To make
these enablers a reality, the intelligent agent described in this paper integrates sensing, learning, and optimization. In order to
meet the current and future needs of 5G and beyond, the article lays out a cross-layer architecture for artificial intelligence (AI)
that is both adaptable and easy to install. We talk about the importance of AI in facilitating network development and show 5G
use cases that are AI-enabled, which incorporate key 5G features.