A Cost Efficient Edge Computing Scheme in Dual-band Cooperative Vehicular Network

With miscellaneous computation-intensive applications generated in vehicular network, e.g., automatic driving and in-car entertainment, it cannot achieve ideal computing performance since vehicles have limited task processing capabilities. To fulfill the goal of low latency and sustainable development in beyond 5G network, we design a cost efficient task processing scheme in dual-band cooperative vehicular network where the tasks can be processed locally, or offloaded to macro-cell base station or road side unit through low-frequency or high-frequency band. Based on this, the sum cost considering energy consumption and latency is minimized through optimizing the task scheduling, computation and communication resource allocation while considering the vehicle sojourn time. Owing to the non-convexity of problem, we decompose it into three subproblems and propose an alternating algorithm to solve them iteratively. Numerical results validate the superiority of the proposed dual-band cooperative task processing scheme and evaluate its performance under different parameter settings, which prove it to be an efficient edge computing scheme for the future vehicular network.