Where to Process: Deadline-aware Online Resource Auction in Mobile Edge Computing

Abstract

Mobile edge computing (MEC) leverages the advan-tages of both cloud computing and mobile computing. An important issue affecting the efficiency of MEC systems is to determine whether letting the mobile users (MUs) offload the computation tasks to the cloudlet or processing the computation tasks locally. The challenge increases in practical MEC systems when the computation tasks have deadlines and the MUs are moving with intermittent wireless local area network (WLAN) connections. To tackle these challenges, in this paper, we propose a Deadlineaware Online Resource Auction (DORA) framework for dynamic computational resource allocation in MEC systems. In DORA, MUs dynamically evaluates the computational resource at the cloudlets and make bidding and offloading decisions. On the other side, the service provider (SP) carries out a winner selection process to make resource allocation and pricing decisions. The proposed DORA framework is truthful and achieves closeto-offline-optimal time-averaged social welfare with polynomial time complexity. Furthermore, the proposed DORA framework introduces a novel method for online resource evaluation in a stochastic setting. Through rigorous theoretical analysis and comprehensive simulations, we demonstrate the effectiveness of the proposed DORA framework.