Security Enhancement of CSI-Based Wireless Sensing via Generative AI
Bibliographic record
Abstract
Integrated Sensing and Communication (ISAC) is becoming a key technology in 6 G networks, where sensing based on channel state information (CSI) plays an essential role. Present research primarily focuses on enhancing sensing performance, yet often overlooks security issue, especially the threat of unauthorized sensing that tends to receive little attention. In response to the above threat, this paper proposes to use generative AI to enhance the security of CSI-based sensing systems. Specifically, we design the guarding signal according to the characteristics of CSI fluctuations caused by user activities and build the corresponding database based on the measurements collected by software-defined radio. Utilizing the constructed dataset, we train the conditional generative diffusion model, which can produce guarding signals that are similar yet distinct from the original training samples. Then, these guarding signals are modulated onto pilot signals, effectively masking the user-induced fluctuations, thereby preventing unauthorized devices from performing illegitimate sensing. Taking the user activity recognition as the example, experimental evaluations illustrate that the proposed method reduces the recognition accuracy of unauthorized devices by about 75 %, significantly enhancing user privacy protection against unauthorized sensing.
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 | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| 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".