Edge-Deployable ML Agent for Real-Time Tactic and Technique Attribution in Microgrid Security
Bibliographic record
Abstract
The proliferation of microgrids increases exposure to advanced persistent threats (APTs) that evolve through structured adversarial stages defined by tactics and techniques. While the MITRE ATT&CK for ICS framework offers a comprehensive threat taxonomy, its integration into edge-level Intrusion Detection Systems (IDSs) remains limited. Existing IDSs mainly perform binary classification (benign vs. malicious) and depend heavily on centralized Security Operations Centers (SOCs) for contextual analysis and response. This work proposes a machine learning-based agent for real-time attribution of adversarial tactics and techniques, leveraging the MITRE ATT&CK framework. Designed as an add-on to existing IDSs, the agent ingests alerts in the standardized Intrusion Detection Message Exchange Format (IDMEF). Its effectiveness is validated through integration with a realistic microgrid environment, supported by real-time experimental data and edge deployment on a Raspberry Pi. A comparative study of classifiers, SVM, Random Forest, Decision Tree, Naive Bayes, and XGBoost has been conducted, with feature selection optimized using Binary Particle Swarm Optimization (BPSO). XGBoost achieved the highest classification accuracy of 98% with an inference latency of 15 ms, and minimal CPU (16.2%) and memory (1.7%) utilization, demonstrating its suitability for resource-constrained edge devices. The results are visualized in Grafana, enabling proactive, tactic-aware defense and improving local situational awareness while reducing depen dency on cloud-based SOCs.
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 | 0.001 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.001 | 0.002 |
| 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".