Autonomous and Adaptive Cyber Incident Detection and Response in Industrial Cyber-Physical Systems Using Hierarchical Reinforcement Learning
Bibliographic record
Abstract
Cyber-Physical Systems (CPSs) are the backbone of many critical infrastructures. However, they have introduced an uncharted territory of security vulnerabilities and attack vectors, mainly due to the deeply integrated physical and cyber spaces. Moreover, in industrial CPS settings, network openness exposes the system to the outside world and renders it vulnerable to cyber threats. The security of industrial CPS significantly relies on the cyber incident detection and response systems which are fundamental to ensure the continuous and proper operation of cyber-physical processes. Among the key configuration parameters of these defense systems is the detection threshold. However, finding the optimal threshold that strikes the right balance between missed detection and false-positive rates remains a challenging problem. In this article, we propose a novel approach that leverages a Hierarchical Reinforcement Learning (HRL) architecture to autonomously detect the dynamic instability in an industrial CPS network and respond by adapting the cyber incident detection and response threshold range to minimize the effects of possible incidents. We developed and tested four HRL algorithmic variants, each offering potential avenues for optimization with its own strengths and limitations. Our agents dynamically select these ranges by assessing the expected risk and potential damage over time. In addition, the agent’s selection process aims to minimize false positives and reduce the cost associated with changing the selected range. All four algorithmic adaptations show the effectiveness of HRL for designing adaptive cyber-physical defense compared to static approaches. Our experimental results indicate that our proposed technique is effective for building autonomous cyber incident detection systems in industrial CPS.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".