A Decentralized Reactive Approach to Online Task Offloading in Mobile Edge Computing Environments

Qinglan Peng, Yunni Xia, Yan Wang, Chunrong Wu, Xin Luo, Jia Lee

Abstract: In mobile edge computing (MEC) environments, the task offloading towards nearby edge servers usually occurs when local resources are inadequate for computation-intensive applications. While the MEC servers benefit from the close proximity to the end-users to provide services at reduced latency and lower energy costs, they suffer from limitations in computational and radio resources, which calls for smart, timely, and efficient offloading methods and strategies. In this paper, we consider an arbitrary request arrival pattern and formulate the MEC-oriented task offloading problem as an online multi-dimensional integer linear programming. We propose a decentralized reactive approach by adopting a dynamic-learning mechanism to yield online offloading decisions upon request arrivals. Experiments based on real-world MEC environment datasets show that our method outperforms state-of-the-art ones in terms of offloading responsiveness and efficiency.