Probabilistic Robustness-driven Universal Adversarial Perturbations with Explainability against Deep Reinforcement Learning-based Intrusion Detection System

2026-09-24

AI Summary

This study addresses the vulnerability of deep reinforcement learning-based intrusion detection systems to universal adversarial perturbation (UAP) attacks and the inadequacy of existing robustness evaluations. To this end, it proposes PX-UAP, a novel method that introduces an explicit probabilistic robustness objective into the UAP generation framework for the first time. Furthermore, PX-UAP leverages explainable artificial intelligence (XAI) to guide perturbation shaping, ensuring that the generated attacks satisfy the constraints of realistic network environments. Experimental results demonstrate that PX-UAP consistently outperforms state-of-the-art methods in attack effectiveness. By integrating probabilistic robustness with XAI-guided perturbation generation, this work establishes a new paradigm for evaluating and enhancing the adversarial robustness of intrusion detection systems against universal adversarial threats.

Abstract

Deep reinforcement learning (DRL) enables adaptive intrusion detection in dynamic network environments but also exposes intrusion detection systems (IDS) to adversarial threats such as universal adversarial perturbations (UAPs), which apply a single input-agnostic perturbation to degrade detection performance across traffic. Probabilistic Robustness (PR), as a post-hoc evaluation metric, provides a principled, population-level measure of adversarial impact that conceptually aligns with the universality objective of UAPs, i.e., PR quantifies the prevalence of misclassification in the input space, making it a natural signal for guiding UAP generation. Hence, we propose PR-based UAP, which represents the first integration of an explicit PR-driven objective into generating UAPs against DRL-based IDS. Building on this formulation, we introduce PX-UAP, which leverages explainable artificial intelligence (XAI) to guide perturbation shaping under realistic domain constraints, and provides a rigorous theoretical analysis of its design. Extensive experiments demonstrate that PX-UAP consistently outperforms state-of-the-art UAP methods in attack effectiveness.

Problem

• Deep Reinforcement Learning
• Intrusion Detection System
• Universal Adversarial Perturbations
• Probabilistic Robustness
• Adversarial Threats

Innovation

• Universal Adversarial Perturbations
• Probabilistic Robustness
• Deep Reinforcement Learning
• Explainable AI
• Intrusion Detection System