Detecting Poisoning Attacks in Collaborative IDSs of Vehicular Networks Using XAI and Shapley Value
Bibliographic record
Abstract
Machine learning-based Intrusion Detection Systems (IDSs) for vehicle networks can collaborate to enhance their performance by sharing crucial decisions when individual datasets lack diversity, which hinders effective model training. However, such collaborative coalitions are vulnerable to poisoning attacks, where certain members tamper with training data, leading to wrong predictions by the IDSs. The current solutions to this problem have the following issues: (1) They require accessing the training dataset of IDSs, which raises critical privacy concerns; (2) their heavy reliance on voting mechanisms may exclude clients and fail to detect malicious coalition members when attackers form the majority coalition; and (3) the verification process can potentially expose sensitive information, posing privacy risks. To address these issues and improve the resilience of collaborative IDSs against poisoning attacks, we propose a novel approach that combines XAI (Explainable AI) technology with Shapley value from cooperative game theory. XAI justifies IDS decisions, while the Shapley value identifies poisoning attacks in training datasets by capturing contradictions between explanations and decisions. We tested our implementation using a publicly available dataset and various machine learning models (e.g., RFC, SVM, and LSTM). Our results prove highly promising in detecting poisoning attacks and overcoming the flaws in existing solutions.
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 | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| 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".