Incentive-Based Distributed Resource Allocation for Task Offloading and Collaborative Computing in MEC-Enabled Networks
Guang Chen,Yueyun Chen,Zhiyuan Mai,Conghui Hao,Meijie Yang,Liping Du
DOI: https://doi.org/10.1109/jiot.2022.3233026
IF: 10.6
2023-05-09
IEEE Internet of Things Journal
Abstract:Computing tasks offloaded from user devices (UDs) can be carried out by one or more mobile edge computing (MEC) servers to alleviate the computing burden of UDs. The incentive is needed to encourage MEC servers to provide their computing services to other network nodes. In this article, inspired by the fact that bargaining games have the available features of incentive, self-enforcement, and satisfaction for all participants, we propose a two-level bargaining-based incentive mechanism for task offloading and collaborative computing in MEC-enabled networks. In the first-level bargaining between UDs and local MEC server (LMECS), both UDs and LMECS try to maximize their respective offloading utilities, which are all defined as a saved-cost function considering the time and energy consumption of task execution, and computing service fees. The task offloading decision, uplink transmitting power of UDs, computing resource allocation of LMECS, and the fees paid by UDs to LMECS are jointly optimized. When large computing tasks are offloaded to LMECS, which results in LMECS overload, the second-level bargaining is proposed to achieve a computing load balance of LMECS and maximize the respective collaboration utilities of LMECS and collaborative MEC server group (CMECG), in which the optimized normalized fees paid by LMECS to CMECG for additional computing resources are obtained. The first-level and the second-level bargainings are proved to be quasi-concave and concave, respectively, and each has a unique Nash bargaining solution (NBS). The simulation results show that the proposed method gets better performance than benchmark methods.
computer science, information systems,telecommunications,engineering, electrical & electronic