DRL for handover in 6G-vehicular networks: A survey

Abstract

3GPP is working on technology improvements related to sixth-generation (6G) wireless communication networks to keep pace. 6G networks are being developed as the next phase forward. Compared to their predecessor wireless technologies, 6G networks are predicted to offer better coverage and flexibility by supporting higher throughput, faster velocities, lower latency, and higher capacity. 6G targets the health, education, industry, and transport sectors. The transport field is undergoing rapid change. 6G and artificial intelligence (AI) will have an essential impact in this area, bringing users new services and functionalities. Within this field, the handover (HO) mechanism remains a concern that researchers must consider for achieving excellent communication quality since HO technology is crucial for ensuring seamless connectivity during user transfers between cells. Numerous proposed Machine Learning (ML) approaches, including Deep Reinforcement Learning (DRL), were discussed to solve HO issues. Recently, DRL methods have garnered significant interest in prospective wireless networks. They can surmount the escalating obstacles of the wireless environment and the constraints of conventional approaches. Moreover, DRL is crucial in wireless networks because of its capability to overcome the specific threats and dynamic nature of wireless settings. Elaborating on a comprehensive survey of these approaches related to HO and DRL can provide a unified analysis of the current advancements. This overview will help to improve understanding of this topic. This survey provides an overview of requirements and usage scenarios for 6G. It highlights the impact of this new wireless technology on the transportation field or the Vehicle-to-everything (V2X). In addition, we provide a deep study of HO management in 6G networks and elaborate on the various DRL literature solutions for HO in mobile and vehicular networks.