Handover management for UAV communication in 5G networks: A systematic literature review

Within the last ten years, UAVs has captured both the academia’s and industry players attention due to their capabilities to be utilized in multiple applications. One of the most prominent functionalities that UAVs provided for multiple applications is communication. Due to these functionalities, UAVs are seen as a major player in the 5G/6G cellular networks. Extending the connectivity and communication through multiple devices is the main goal of the 5G/6G networks. In this context, UAVs can either be deployed as relays that receive signals from BS and transmit them to the UE, or the UAVs can be utilized as a BS that flies (UAV-BS) and supplies 5G/6G communication to the UEs on the ground. However, there are many issues that arise with UAV communication networks when trying to provide signal coverage to users. One of the issues is when utilizing UAV as a BS, the signal coverage’s quality might not be up to par due to the UAV’s high mobility characteristics, which also leads to the frequent handover experienced by the users on the ground. Frequent handovers, or “ping-pong” handovers, are not acceptable and can instigate other problems that reduce the quality of the signal, such as packet delays, or losses of packets. Efficient handover management in drone communication is the solution to this problem, and, thus, it requires to be applied to the UAV networks to sustain connectivity that is reliable and stable. This paper studies the handover management for UAV communication in 5G networks using the systematic literature review (SLR) technique for its methodology. Multiple topics related to the handover scenario in the UAV networks are being highlighted in this paper, alongside studies related to machine learning (ML) usage in making handover decisions. A total of 90 research papers that relate to the handover in UAV networks and that were published from 2019 to 2024 were chosen. The results from our studies show that machine learning is able to provide handover decision that are more efficient and robust than traditional/conventional methods of making handover decisions. Finally, several discussions based on several themes and the limitations of the studies were held.