Budget-Feasible Truthfulness Mechanism for Task Offloading and Interaction in Edge-Vehicle Collaborative Computing

Mobile-edge computing (MEC) affords high-computing power but lacks sensing capability. Furthermore, intelligent vehicles, which possess rich sensing resources, consume limited energy. Motivated by this, we propose edge-vehicle collaborative computing and investigate the task offloading and interaction problem (TOIP), in which MEC servers and vehicles collaborate to leverage their strengths and mitigate their weaknesses. Motivated by practical application requirements, we propose a task interaction model where a user’s computing and sensing subtasks are, respectively, offloaded onto the MEC servers and vehicles, which then collaborate to complete the tasks. Aiming to maximize group efficiency, we formulate the TOIP in an auction-based setting. To motivate MEC servers and vehicles, we propose a reverse auction where each user is an auctioneer, while MEC servers and vehicles are the bidders. Our reverse auction mechanism achieves budget feasibility, where the rewards received by MEC servers and vehicles cannot exceed the budget. The proposed mechanism proves to be truthful; that is, MEC servers or vehicles cannot obtain higher utility by declaring untrue values. We also demonstrate how to make the mechanism meet the truthfulness requirement in TOIP. In addition, the proposed mechanism achieves individual rationality, consumer sovereignty, and computation efficiency. We also theoretically analyze the approximate ratio. The simulation results show that the proposed mechanism exhibits exceptional performance in all the scenarios.