Federated Secure Intelligent Intrusion Detection and Mitigation Framework for SD-IoT Networks using ViT-GraphSAGE and Automated Attack Reporting
Bibliographic record
Abstract
Software-Defined Internet of Things (SD-IoT) networks enabled intelligent network management through their dynamic features but expose centralized infrastructure to complex cyberattacks that put the system in great danger. In order to address this, a novel Federated Secure Intelligent Intrusion Detection and Mitigation framework with Automated Attack Reporting for SD-IoT network is proposed. The combination of Vision Transformer (ViT) and GraphSAGE architecture enables the model to process network relationships globally and locally which effectively increases intrusion detection performance. Besides, in real-time, to dynamically reroute network traffic and isolate the compromised nodes, a Multi-Agent Deep Q-Learning (MA-DQL) based mitigation strategy is employed which minimizes attack impact. For enabling collaborative and secure communication without centralized data exposure, the edge nodes are integrated with Federated Learning (FL) that ensures privacy-preserving and distributed model training. The proposed system also incorporates a Flan-T5 Transformer-based Automated Attack Reporting System which develops comprehensive forensic reports that expose security threats and their corrective actions. Through this proposed framework, accurate threat detection of 98.06% with real-time adaptative operation and efficient attack containment is accomplished in parallel with reduced computational load which ensures secure operation of SD-IoT systems.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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 source (direct Gemma or distilled Codex), 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".