Scholarlink

This paper presents a dynamic defense strategy against intelligent agent-based attacks. It employs deceptive techniques—such as altering sensitive host addresses and obfuscating operating system information—to mislead adversaries. A situation-awareness module is integrated to adjust defense actions in real time based on network topology metrics. Compared with conventional honeypot defenses, the proposed situation-aware moving target defense demonstrates superior performance in mitigating agent-driven threats. To address the high adaptability and stealth of AI-based attacks, as well as the overhead of MTD, the strategy leverages Netfilter and eBPF for efficient OS-level obfuscation. Experiments on the NASim platform confirm its effectiveness in enhancing defensive performance against intelligent agents.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Multi-stage Proactive Defense Against Intelligent Penetration Attacks in Power Systems

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Shiying Feng,
  • Hongyue Ma

摘要

This paper presents a dynamic defense strategy against intelligent agent-based attacks. It employs deceptive techniques—such as altering sensitive host addresses and obfuscating operating system information—to mislead adversaries. A situation-awareness module is integrated to adjust defense actions in real time based on network topology metrics. Compared with conventional honeypot defenses, the proposed situation-aware moving target defense demonstrates superior performance in mitigating agent-driven threats. To address the high adaptability and stealth of AI-based attacks, as well as the overhead of MTD, the strategy leverages Netfilter and eBPF for efficient OS-level obfuscation. Experiments on the NASim platform confirm its effectiveness in enhancing defensive performance against intelligent agents.