MEC-assisted Vertical Handover Based on Improved DQN in Heterogeneous Internet of Vehicles

Vertical handover in heterogeneous networks is a critical and much-discussed topic within the field of mobile communications. It primarily denotes the procedure by which the user transitions from its current network connection to a different type of network, especially in the multifaceted context of heterogeneous vehicular networks. Given the scenarios marked by high mobility, network diversity, and an array of service types, orchestrating a handover that achieves both low latency and high reliability presents a considerable challenge. The endeavor to refine vertical handover processes is pivotal for enhancing user experience and ensuring seamless service continuity in these complex network environments. In this study, we introduce an enhanced handover decision algorithm based on Deep Q-Network (DQN), which is implemented on Mobile Edge Computing (MEC) servers strategically positioned near base stations. We leverage the low-latency characteristics of MEC and its contextual information advantages over wireless communication systems. Our methodology takes into account various critical factors, including the type of terminal services, network parameters on the network side, and the costs associated with handovers. We meticulously craft reward functions for the algorithm, which distinctly categorize handovers as either essential or opportunistic, thereby optimizing the handover decision-making process. Experimental results demonstrate that our algorithm significantly enhances Quality of Service (QoS) and reduces handover latency.