ADVERSARIAL JAMMING FOR AUTOENCODER DISTRIBUTION MATCHING
Waleed El-Geresy, Deniz Gündüz · 2025
Read ADVERSARIAL JAMMING FOR AUTOENCODER DISTRIBUTION MATCHING by Waleed El-Geresy, Deniz Gündüz (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-02740.txt.
Description
Waleed El-Geresy and Deniz Gündüz turn an attacker into a training tool for an autoencoder. An autoencoder squeezes data into a small latent code, then rebuilds the original from it. For the code to behave well, its spread should match a chosen bell shaped distribution.
A wireless jammer tries to disrupt recovery of a Gaussian source sent over the channel. Fighting that jamming pushes the encoder to shape its latent space as desired. The setup joins learning with adversarial pressure, a trick that may help design robust communication codes. Noise is most often the foe in radio work, so a scheme that puts noise to work stands out from the pack.
Catalog record
| CREATOR | Waleed El-Geresy, Deniz Gündüz |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 19 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2512-02740