Deep Reinforcement Learning Based Access Control Strategy for Edge Computing in IoT System
Zhihao. Yu, Wanda. Chen, Jie. Wang, Kaiwen. Ye · 2021 IEEE International Conference on Computer Science, Electronic Information Engineering and Intelligent Control Technology (CEI) · 2021
As edge computing has been widely used in IoT (Internet of Things) systems, the multi-service access control has become one of important issues for IoT. That is because current edge computing technologies fail to provide supports to fine-grained multiple services with delay sensitivity and real time demands. To solve this problem, this paper proposes a deep reinforcement learning based access control scheme of the edge computing device in IoT system, in which a novel edge computing network architecture is build. by making full use of deep reinforcement learning algorithm, this proposed strategy is also designed to perform flexibly resources allocation for fine-grained services. Testing results show the feasibility and reasonability of the proposed scheme with better resources utilization under limited conditions.