Lightweight Security Mechanism Based on Heterogeneous Construction for Virtualized Edge System: A Markov Decision Process Approach
Qimei Cui, Changzheng Huang, Zengbao Zhu, Xueqing Huang, Na Li, Xiaofeng Tao, Ping Zhang · 2020
Security issues play a crucial role in the development of 5G networks and beyond. By deploying computing resources at the access networks, the mobile edge computing paradigm in 5G networks can provide users with high-quality services in terms of low latency and high throughput. However, virtualized edge components, limited edge cloud resources, and proximity with the user plane bring a larger attack surface to the edge systems. The lightweight proactive defense mechanism is a promising approach to address rising security challenges. In this paper, a novel security structure based on heterogeneous construction is proposed to deal with a variety of potential security threats to the virtualized edge system. The proposed lightweight security mechanism maintains abnormal components in the structure flexibly through discontinuous detection, ensuring the security and usability of the system with low maintenance overhead. To optimize the long-term utility of our structure in terms of effective throughput, task failures, and maintenance costs, the Markov Decision Process (MDP) method is adopted to construct the utility of the proposed structure as the reward function. The optimal policy of each state is obtained via a value iteration algorithm. The simulation results show that the proposed structure and optimal policy achieve less breakdown rate and better utility, as compared to the heuristic policies.