QoS‐Aware Vertical Handover Management in Internet of Vehicles Using a Hybrid Optimization Framework

The Internet of Vehicles (IoV) connects to the internet via a 5G network and covers a wide range of different network architectures. In this environment, ubiquitous network connections are formed. High mobility of vehicles requires frequent handovers in heterogeneous networks due to the presence of multiple radio access technologies. For vehicles at short range, dedicated short‐range communication (DSRC) is used, while better technology is needed for long‐range communication. 4G long‐term evolution (4G LTE) features large bandwidth and wider cell coverage but has high transmission time intervals. 5G provides higher data rates and reduced transmission latency; but because it cannot penetrate thick objects, it often suffers from blockage issues. The integration of DSRC and cellular network technology in IoV has aroused research interest in recent years due to the need to study and identify appropriate network selection in a heterogeneous IoV (HIoV) environment so that handover can be done efficiently. Previous research selects a network directly or connects to available radio access networks, resulting in frequent handovers. To address this challenge, a hybrid algorithm in heterogeneous IoV network (HAHIN) is proposed as an integrated framework that conditionally coordinates handover decision, network selection, and vehicle‐to‐vehicle (V2V) route optimization. The proposed work develops an integrated decision framework for HIoV by coordinating handover validation, network selection, and V2V routing across DSRC, 4G LTE, and 5G mmWave technologies, where handover necessity is explicitly validated before initiating target‐network selection and subsequent V2V route optimization. The necessity for a handover is determined using a dynamic Q‐learning algorithm. Network selection is performed through a fuzzy convolutional neural network (F‐CNN) algorithm. When data transfer is required between vehicles, the jellyfish search optimization (JSO) algorithm facilitates V2V chain routing. Thus, the three established techniques are applied sequentially to address distinct but interconnected HIoV decision stages. The validation results and performance evaluations demonstrate that the proposed approach outperforms existing state‐of‐the‐art methods.