mlsanityData to 7 Oct 2026

Deep PackGen: A Deep Reinforcement Learning Framework for Adversarial Network Packet Generation

arXiv 2305.11039 ↗First posted 18 May 2023cs.CR cs.LGcited by 5

Not on the map. Why?

Abstract (arXiv)

Recent advancements in artificial intelligence (AI) and machine learning (ML) algorithms, coupled with the availability of faster computing infrastructure, have enhanced the security posture of cybersecurity operations centers (defenders) through the development of ML-aided network intrusion detection systems (NIDS). Concurrently, the abilities of adversaries to evade security have also increased with the support of AI/ML models. Therefore, defenders need to proactively prepare for evasion attacks that exploit the detection mechanisms of NIDS. Recent studies have found that the perturbation of flow-based and packet-based features can deceive ML models, but these approaches have limitations. Perturbations made to the flow-based features are difficult to reverse-engineer, while samples generated with perturbations to the packet-based features are not playable. Our methodological framework, Deep PackGen, employs deep reinforcement learning to generate adversarial packets and aims to overcome the limitations of approaches in the literature. By taking raw malicious network packets as inputs and systematically making perturbations on them, Deep PackGen camouflages them as benign packets while still maintaining their functionality. In our experiments, using publicly available data, Deep PackGen achieved an average adversarial success rate of 66.4% against various ML models and across different attack types. Our investigation also revealed that more than 45% of the successful adversarial samples were out-of-distribution packets that evaded the decision boundaries of the classifiers. The knowledge gained from our study on the adversary's ability to make specific evasive perturbations to different types of malicious packets can help defenders enhance the robustness of their NIDS against evolving adversarial attacks.

Key facts. Cited by 5 papers in the map (data through 7 Oct 2026). Builds on 18 of its 64 references that are in the map. First posted on arXiv 18 May 2023.

Builds on

18 of 64 references.

References · 18

  1. 2017A Unified Approach to Interpreting Model Predictionscited by 6,488
  2. 2013Intriguing properties of neural networkscited by 5,114
  3. 2015Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classificationcited by 3,590
  4. 2015Deep Reinforcement Learning with Double Q-learningcited by 1,704
  5. 2017Deep Packet: A Novel Approach For Encrypted Traffic Classification Using Deep Learningcited by 69
  6. 2020Adversarial Machine Learning in Network Intrusion Detection Systemscited by 25
  7. 2020Evaluating and Improving Adversarial Robustness of Machine Learning-Based Network Intrusion Detectorscited by 25
  8. 2019FENCE: Feasible Evasion Attacks on Neural Networks in Constrained Environmentscited by 23
  9. 2018IoT Security Techniques Based on Machine Learningcited by 22
  10. 2020UAV Coverage Path Planning under Varying Power Constraints using Deep Reinforcement Learningcited by 15
  11. 2020Adversarial Network Traffic: Towards Evaluating the Robustness of Deep Learning-Based Network Traffic Classificationcited by 11
  12. 2021TANTRA: Timing-Based Adversarial Network Traffic Reshaping Attackcited by 11
6 more
  1. 2020Deceiving Image-to-Image Translation Networks for Autonomous Driving with Adversarial Perturbationscited by 9
  2. 2020A Reinforcement Learning Approach for Rebalancing Electric Vehicle Sharing Systemscited by 7
  3. 2018Improving Network Intrusion Detection Classifiers by Non-payload-Based Exploit-Independent Obfuscations: An Adversarial Approachcited by 4
  4. 2021Packet-Level Adversarial Network Traffic Crafting using Sequence Generative Adversarial Networkscited by 3
  5. 2019Knowledge-Based Sequential Decision-Making Under Uncertaintycited by 2
  6. 2022Deep VULMAN: A Deep Reinforcement Learning-Enabled Cyber Vulnerability Management Frameworkcited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 1; 2025: 2; 2026: 2.

5 papers.

  1. 2025Privacy-R1: Privacy-Aware Multi-LLM Agent Collaboration via Reinforcement Learningcited by 6
  2. 2026SoK: The Pitfalls of Deep Reinforcement Learning for Cybersecuritycited by 4
  3. 2025Adversarial Agents: Black-Box Evasion Attacks with Reinforcement Learningcited by 3
  4. 2026Building Better Environments for Autonomous Cyber Defencecited by 3
  5. 2024A Synergistic Approach In Network Intrusion Detection By Neurosymbolic AIcited by 2