Generative AI-Empowered Resilient Adaptive ISAC Against Adversarial Machine Learning Attacks
Bibliographic record
Abstract
In this paper, we present a novel resilient adaptive integrated sensing and communication (RAd-ISAC) framework. It aims to enhance ISAC systems for advanced driver-assistance systems (ADASs) to mitigate adversarial machine learning (AML) threats and improve resource efficiency. AML attacks target ADASs by injecting false Range-Doppler maps (RDMs) to mislead the system about target vehicles (TVs), resulting in resource inefficiency. Our approach introduces a generative adversarial network (GAN) equipped with a differentiable Kolmogorov-Smirnov (KS) loss function, termed KSGAN. This significantly enhances AML attack detection by generating highly realistic RDM samples, improving the robustness of the AML detector. To optimize resource allocation in ISAC systems, we propose an adaptive signal transmission method. This allows the source vehicle (SV) to switch dynamically between ISAC and communication-only signals based on the AML detector's output and 2D constant false alarm rate (CFAR) analyses. We conduct extensive simulations with synthetic data using IBM's adversarial robustness toolbox (ART). Our results show that KSGAN outperforms standard GAN, Wasserstein GAN (WGAN), and relational GAN (RGAN) in AML detection. Additionally, when compared to other ISAC designs, including standalone ISAC, Faster-than-Nyquist ISAC (FTN-ISAC), ISAC-accelerated edge intelligence system, and hybrid-ISAC, our Rad-ISAC framework achieves the lowest root mean square error (RMSE) and Cramér-Rao lower bound (CRLB). This work highlights an unexplored vulnerability of ISAC systems to AML attacks and demonstrates advancements in ADAS vehicle safety and efficiency.
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 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 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 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 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".