Digital-Twin-Empowered Interference Management for Multihop Internet of Vehicles Networks Over Millimeter Wave Bands

The Internet of Vehicles (IoV) generates massive data traffic and demands reliable end-to-end connectivity to achieve multi-Gbps throughput between vehicles and roadside units. Millimeter-wave (mmWave) bands, with their abundant bandwidth, are promising for high-throughput IoV networks. However, in this context, significant propagation losses, intermittent line-of-sight availability, and dynamic topology changes due to vehicle mobility present critical challenges. This article introduces resource allocation for vehicular networks ( $\textsf {RAVEN}$ ), a centralized resource management framework designed to address these challenges effectively. $\textsf {RAVEN}$ leverages a digital twin network (DTN) to optimize the end-to-end system capacity of multihop mmWave IoV networks by effectively managing co-channel interference among vehicles. $\textsf {RAVEN}$ comprises the following three steps: 1) a channel prediction step that utilizes DTN’s awareness of vehicular mobility and environmental contexts to predict site-specific channel gains for vehicular communication links; 2) a clustering step that partitions vehicles into nonoverlapping clusters, allowing vehicles within each cluster to share the same mmWave channel for data transmission, while simultaneously reducing co-channel interference; and 3) a multihop connectivity optimization step that provides a connected vehicular networking topology by jointly optimizing vehicle-to-vehicle and vehicle-to-infrastructure connectivity using a graph theory approach. A proof-of-concept of $\textsf {RAVEN}$ is developed by implementing a DTN on the Microsoft Azure Digital Twins platform while integrating real-world vehicular mobility traces, edge-cloud collaboration, and parallel computing. Extensive simulations demonstrate that $\textsf {RAVEN}$ outperforms several benchmark schemes, and offers scalability and near real-time decision-making capabilities for managing interference in large-scale IoV networks.