Optimization Design of Network Attack and Defense Scenarios in Intelligent Clusters
Chengyue Wang, Yang Liu, Yaojun Qiao, Daoqi Han, Yueming Lu · 2024
Intelligent cluster networks are crucial infrastructures with a high susceptibility to cyber threats, necessitating the optimization of existing security protocols and resource allocation schemes to dynamically mitigate these risks. This optimization serves as the cornerstone for next-generation security measures. In the realm of defense strategy research, accurately depicting network attack and defense scenarios is pivotal. However, much of the existing research in this domain primarily focuses on theoretical scenarios, lacking a thorough analysis of real-world characteristics, thereby hindering the efficacy of algorithmic applications. To address this gap, we utilize incomplete information and zero-sum game theory to establish and refine attack and defense models within intelligent cluster networks. Leveraging these models, a game simulation system is constructed using the Gym toolkit to enhance the realism of network security simulation scenarios. Additionally, we design six typical attack strategies to more precisely simulate common attacks within intelligent networks, thereby facilitating the development of novel defensive strategies. Experimental results demonstrate that the proposed network attack and defense simulation model is well-designed, and the performance of the attack strategies closely mirrors real-world attacks. This provides a simulation environment closely resembling actual scenarios, thereby offering a robust platform for the exploration of defense strategies within intelligent cluster networks.