Resource-Aware Multi-Hop Routing Protocol for Unicast Cellular V2V Communications

Using a multi-hop routing protocol for unicast vehicle-to-vehicle (V2V) communications is crucial to facilitate data relaying from one vehicle to a distant one. The dynamic behavior of vehicular ad-hoc networks (VANETs) and unstable relays often cause frequent disconnections or switchovers, leading to higher latency, diminished reliability, and increased resource use. The stability of these routes depends not just on link connectivity, but also on the availability of adequate resources. Although numerous studies have explored traditional VANET routing protocols to tackle these issues, they often neglect the critical aspect of resource availability. In this study, we focus on a V2V routing method that considers resource availability, assuming the use of a geo-based resource allocation framework. In the proposed resource-aware approach, we utilize deep learning to predict vehicle trajectories and traffic load in various areas. The proposed resource-aware routing protocol aims at improving both spectrum efficiency and route stability by employing distance between vehicles, traffic load, and resource availability of areas to manage emergency messages (EMs) more efficiently and reduce unnecessary rebroadcasting and congestion. Our simulation results reveal that our proposed method surpasses traditional protocols in terms of packet reception ratio (PRR), latency, and average hop count.