Review of Artificial Intelligence Based Beam Tracking Techniques for mmWave 5G and Beyond Networks

Abstract

Wireless networks of the future can take advantage of beamforming techniques in the millimeter wave (mmWave) and terahertz (THz) bands to effectively handle the immense bandwidths required. This opens up a world of possibilities for the advancement of wireless technology and the potential to create even faster and more efficient networks. To achieve directional beamforming gain, it is essential to have a reliable beam management (BM) framework that can track the best uplink and downlink beam pairs using traditional exhaustive beam scans (EBS). However, this requires extensive beam measurement, which can result in a significant overhead, especially for higher carrier frequencies and narrower beams. To tackle this issue, machine learning (ML) algorithms based on artificial intelligence (AI) are being created to detect and understand intricate mobility patterns and environmental changes. This article presents an overview of the current AIbased ML beam tracking (BT) techniques used in mmWave/THz bands for 5G and future networks, highlighting the essential features of an effective beam tracking framework.