Digital twin-assisted resource allocation framework based on edge collaboration for vehicular edge computing

Sekione Reward Jeremiah, Laurence Tianruo Yang, Jong Hyuk Park

Abstract: Highlights•This study investigates the edge collaboration problem, resource allocation, and RSUs association (selection) in a DT-assisted VEC system with NOMA for vehicles. First, we discussed the VEC network architecture and identified key research contributions concerning vehicular communication and resource allocation. We then presented the DT concept alongside our proposed DT-assisted VEC network framework for edge collaboration and resource allocation.•We aimed to improve the VEC network performance by minimizing total task completion delays and maximizing system-wide throughput by jointly optimizing vehicle offloading decisions, RSU’s association, and subchannel allocation for vehicles.•The digital twin creates a virtual representation of physical networks to predict, estimate, and analyze the real-time network state. A digital twin network (DTN) is also formed to manage the physical network easily.•The proposed DT-based VEC network assists in achieving adaptive network orchestration through the help of the A2C DRL model for resource management and network state estimation.•The integration of DT and VEC is elaborated by solving the offloading and resource allocation problem using the DRL model, where virtual models of vehicles and RSU are built for that purpose.