Projected Natural Gradient Method: Unveiling Low-Power Perturbation Vulnerabilities in Deep-Learning-Based Automatic Modulation Classification
Bibliographic record
Abstract
Rapid advancements in deep learning (DL) and the availability of the large data sets have made the adoption of DL highly appealing across various fields. Wireless communication systems, including future 6G systems are anticipated to incorporate intelligent components like automatic modulation classification (AMC) for the cognitive radio and dynamic spectrum access. However, DL-based AMC models are susceptible to the adversarial attacks, which consist of crafted perturbations that aim to alternate the decision of a victim model. This study focuses on investigating and uncovering modern modulation classifiers’ vulnerability to the adversarial threats. Though attacks of this nature inherently jeopardize DL-based classifiers, contemporary attack methods typically exhibit diminished impact at the lower perturbation levels. Therefore, we introduce a novel attack approach that exploits the Riemannian manifold properties of the intricate neural networks, yielding adversarial samples with heightened efficacy at the lower perturbation powers. We thoroughly evaluate how effective various defense techniques are and demonstrate our proposed attack method’s ability to thwart them. The findings of this study shed light on the limitations and vulnerabilities of the DL-based AMC models in the face of the adversarial attacks. By addressing these challenges, we can enhance the robustness and security of these models, and pave the way for their reliable deployment in practical wireless communication systems, including the future 6G 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 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 | 0.002 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| 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 | 0.002 |
| 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".