Defending Against APT Attacks in Cloud Computing Environments Using Grouped Multiagent Deep Reinforcement Learning
- 21 February 2025
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (12) , 19459-19470
Abstract
Advanced Persistent Threats (APTs) pose a significant challenge to cloud computing security in the evolving landscape of cyber threats. Traditional defense models rely heavily on the attacker’s historical attack information, which greatly limits the effectiveness of actually dealing with APT attacks. To address this issues, we investigate an attack-defense game model in clouding computing environments, where multiple attackers and multiple defenders are supposed to compete for resource allocation on the cloud servers. In order to develop more effective defense strategies, we formulate the optimization problem to maximize the average rewards of defenders under constraints of the maximum available resource and acceptable cost. To solve this, we propose to use the multi-agent deep reinforcement learning method to cope with the high uncertainty and dynamics of attack behavior. Then it is proposed to divide all defenders into cooperative groups and allow defenders within each group can jointly optimize the defense strategy through sharing information and experience. On this basis, we propose a novel grouped multi-agent deep reinforcement learning defense (GMADRLD) algorithm, which can effectively mitigate the issue of state space explosion while achieving good defense effect. Simulation results not only demonstrate the effectiveness of the proposed GMADRLD algorithm in dealing with the attacker’s ever-changing strategies, but also show that it is able to strike a balance between defense performance and computational complexity.Keywords
Funding Information
- National Key Research and Development Program of China (2020YFB1805400)
- National Natural Science Foundation of China (62101358, 62402330)
- China Postdoctoral Science Foundation (2020M683345, 2024M752211)
- Sichuan University-Dazhou People’s Government Strategic Cooperation Special Fund Project (2022CDDZ-10)
- Fundamental Research Funds for the Central Universities (2022CDDZ-10-SCU)
- Youth Science Foundation of Sichuan (2025ZNSFSC1474)
This publication has 38 references indexed in Scilit:
- Explainable deep learning approach for advanced persistent threats (APTs) detection in cybersecurity: a reviewArtificial Intelligence Review, 2024
- Stochastic Dynamic Information Flow Tracking game using supervised learning for detecting advanced persistent threatsAutomatica, 2023
- Game-theoretic APT defense: An experimental study on roboticsComputers & Security, 2023
- Analyzing Advanced Persistent Threats Using Game Theory: A Critical Literature Review2022
- Reinforcement learning for combinatorial optimization: A surveyComputers & Operations Research, 2021
- A novel approach for APT attack detection based on combined deep learning modelNeural Computing and Applications, 2021
- Recent security challenges in cloud computingComputers and Electrical Engineering, 2018
- A survey of game theoretic approach for adversarial machine learningWIREs Data Mining and Knowledge Discovery, 2018
- A comprehensive analysis for fair probability marking based traceback approach in WSNsSecurity and Communication Networks, 2016
- An empirical comparison of botnet detection methodsComputers & Security, 2014