Extensive Edge Intelligence for Future Vehicular Networks in 6G

Weijing Qi,Qian Li,Qingyang Song,Lei Guo,Abbas Jamalipour

DOI: https://doi.org/10.1109/mwc.001.2000393

IF: 12.777

2021-08-01

IEEE Wireless Communications

Abstract:The 6th generation mobile network (6G) is expected to achieve a fully connected world. As a key enabling technology in 6G, edge intelligence (EI) combines artificial intelligence (AI) with mobile edge computing (MEC), further releasing the potential of edge-side data. However, edge resources from roadside units (RSUs) and base stations are insufficient to match the wide variety of services in future vehicular networks. Extensive EI with auxiliary edge resources from parked vehicles is therefore introduced in this article. To achieve the coordination between extensive edge resources and EI-assisted services for vehicular networks in 6G, we propose an intelligent service-ori-ented edge resource management architecture, which makes full use of knowledge extracted from a large amount of sensory data in vehicular networks. As a case study, a prediction-based edge resource activation method under this architecture is proposed, where resources of parked vehicles are activated based on the predicted user density. Simulation results demonstrate that our prediction model has high accuracy and the pre-diction-based resource activation approach can reduce deployment cost.

computer science, information systems,telecommunications,engineering, electrical & electronic, hardware & architecture