Efficient Adversarial Attacks Against DRL-Based Resource Allocation in Intelligent O-RAN for V2X
Bibliographic record
Abstract
Artificial intelligence (AI) is projected to be a critical part of open radio access networks (O-RAN) to enable intelligence for connectivity management in smart vehicle-to-everything (V2X) networks and vehicle road cooperation systems. However, the openness and dependence of AI models on massive volumes of data render them subject to serious security vulnerabilities, such as adversarial attacks. This study investigates security issues in O-RAN's near real-time RAN intelligent controller (RIC), with an emphasis on deep reinforcement learning (DRL)-based resource allocation. We introduce a novel attack manipulating environmental observations to mislead AI agents, resulting in erroneous allocations and decreased physical resource block (PRB) transmission rates for various vehicular communications. We also discover flaws where compromised users or signal jammers can fake signal power to trick the AI agent's state observation. This can lead to a policy infiltration attack that makes the network performance drop significantly. Evaluation results show up to a 40% decline in user data rates, a 77.74% reduction in packet delivery rates, and significant disruptions in ultra-reliable and low-latency communications (uRLLC) services such as remote driving and connected automated vehicles. The policy infiltration attack causes a 20% increase in packet losses and up to 150% delay overall. The attack efficiency emphasizes the need for adversarial training in protecting AI-driven applications, which should be addressed in future O-RAN security specifications and AI-powered vehicular networks.
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) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| 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".