人工智能驱动的智能蜜罐体系架构与关键技术研究

Research on the Architecture and Key Technologies of AI-Driven Smart Honeypot Systems

摘要: 为了提升蜜罐系统应对高级持续性威胁的智能化水平，本文构建了一种人工智能驱动的智能蜜罐体系架构，研究内容涵盖蜜网场景智能生成、蜜网运维智能响应、威胁数据智能分析推理与溯源等关键技术方向，方法上结合大语言模型、图神经网络与强化学习技术，并引入ATT&CK框架对攻击行为进行建模。研究结果表明，该架构可显著增强蜜罐系统的拟真性、自主响应能力与威胁识别效率。研究具有推动蜜罐从被动诱导向主动智能演进的现实意义，为网络空间主动防御提供了技术支撑与理论依据。

Abstract: To enhance the intelligence of honeypot systems in countering advanced persistent threats, this paper constructs an AI-driven intelligent honeypot architecture. The research covers key technical directions such as intelligent generation of honeynet scenarios, intelligent operational response of honeynets, intelligent analysis and reasoning of threat data, and traceability. Methodologically, it integrates large language models, graph neural networks, and reinforcement learning techniques, while introducing the ATT&CK framework to model attack behaviors. The results demonstrate that this architecture significantly improves the realism, autonomous response capability, and threat identification efficiency of honeypot systems. The study holds practical significance in advancing honeypots from passive detection to active intelligence, providing technical support and theoretical foundations for proactive cyber defense.