Novel method of mobile edge computation offloading based on evolutionary game strategy for IoT devices

• Published in: International journal of electronics and communications Vol. 118; p. 153134
• Main Authors: Cui, Yuya, Zhang, Degan, Zhang, Ting, Chen, Lu, Piao, Mingjie, Zhu, Haoli
• Format: Journal Article
• Language: English
• Published: Elsevier GmbH 01.05.2020
• Subjects: Computation offloading; Evolutionary game theory; Internet of Things; Mobile edge computing; Reinforcement learning; Internet of Things; Computation offloading; Mobile edge computing; Reinforcement learning; Evolutionary game theory
• ISSN: 1434-8411
• DOI: 10.1016/j.aeue.2020.153134

Due to the limited computing resources and energy of IoT devices, complex computing tasks are offloaded to sufficient computing servers, such as Cloud Center. However, offloading may increase the latency and congestion of the IoT network. Mobile Edge Computing (MEC) is a promising approach, which can decrease the delay and energy consumption of IoT devices significantly. In this paper, we investigate the problem of multi-user computation offloading under dynamic environment. Considering the channel interference when multiple IoT devices offload computing tasks via wireless channels at the same time, we formulate the computation offloading as an evolutionary game model. We use the replicator dynamics to analyze the evolutionary process of IoT devices and prove that multi-user computation offloading exists unique Evolutionary Stability Strategy (ESS). Finally, we design an evolutionary game algorithm based on reinforcement learning in practical application scenarios. Experiments can verify the convergence and performance of the proposed algorithm in multi-user scenarios.