Application of deep reinforcement learning for intrusion detection in Internet of Things: A systematic review

2025-02-01

Internet of Things

AI Summary

To address the insufficient adaptability of intrusion detection systems (IDS) in Internet-of-Things (IoT) environments under dynamic threats, this paper presents a systematic review of deep reinforcement learning (DRL) for IoT intrusion detection. It introduces, for the first time, a taxonomy of DRL-based IDS methods specifically tailored to IoT scenarios, uncovering critical research gaps—particularly concerning state representation sparsity and real-time constraints. Empirical evaluations are conducted across heterogeneous datasets (NSL-KDD, CIC-IDS2017, and IoT-specific benchmarks) using representative DRL algorithms including DQN, PPO, and A3C, yielding performance boundaries for 12 canonical method categories. Furthermore, we propose a three-dimensional evaluation framework—encompassing deployability, generalizability, and communication overhead—to rigorously assess practical viability. This work provides both theoretical foundations and actionable guidelines for developing lightweight, adaptive IDS solutions for resource-constrained IoT deployments.

Technology Category

• Application Domains: Internet of Things, Sensor Networks & Smart Cities
• Knowledge Representation and Reasoning: Description Logics
• Machine Learning: Reinforcement Learning

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Problem

• Enhancing IoT security using Deep Reinforcement Learning
• Improving adaptive threat detection in dynamic IoT networks
• Addressing research gaps in DRL-based Intrusion Detection Systems

Innovation

• Deep Reinforcement Learning enhances IDS adaptability
• DRL improves threat detection and reduces false positives
• Systematic review identifies gaps for future IoT security