A near-optimal resource allocation strategy for minimizing the worse-case impact of malicious attacks on cloud networks

Source: Journal of Cloud Computing, vol. 14, no. 1, art. 41 (2025)

Authors: Yu-Fang Chen, Frank Yeong-Sung Lin, Kuang-Yen Tai, Chiu-Han Hsiao, Wei-Hsin Wang, Ming-Chi Tsai, Tzu-Lung Sun

URL: https://doi.org/10.1186/s13677-025-00749-6

Abstract: The cloud industry has evolved significantly, driven by advancements in network infrastructure and business applications. However, security concerns, especially attack-defense scenarios related to
hacker computing attacks, remain a critical challenge. Despite existing detection systems, sophisticated attacks continue to evade identification. This study presents a novel, optimization-based resource allocation strategy designed to mitigate the worst-case impacts of hacker computing attacks in cloud computing centers. The proposed model integrates Virtual Machine (VM) initiation decisions and employs the Contest Success Function (CSF) within a two-player max-min game framework to dynamically allocate resources. This approach effectively balances security, cost, and service quality. The model not only enhances defense mechanisms against attacks but also optimizes resource utilization, reducing operational costs by 25% while maintaining high levels of security and improving resource efficiency by 30%. Its dual-solution methodology ensures scalability, making it applicable to both small and large-scale cloud environments. The innovative integration of economic theory and multi-objective optimization offers cloud service providers a powerful tool to enhance reliability, security, and cost-effectiveness.