Byzantine attack on AI relay for semantic communications
Original title: AI-Native Task-Oriented Semantic Communications over a Relay: Benefits and Vulnerabilities
Relay-assisted semantic comm improves inference accuracy in low-SNR regimes but relay can inject learned perturbations to degrade destination accuracy.
Semantic communication over a relay node conveys task-relevant classification labels for images. Joint end-to-end design optimizes semantic accuracy and perceptual quality. Relay outperforms direct transmission especially at low SNR. However relay injects structured perturbations via learned mapping, significantly degrading inference.
- Gains most pronounced in low SNR regimes
- Attack injects perturbations on intermediate representations
- Learned mapping applied at test time
- Clear trade-off between benefits and insider security risks
Impact: degraded semantic inference accuracy
From the original article
1 min read at arxiv.orgSemantic communication (SemCom) using a relay node is considered, where the relay assists in conveying task-relevant information from a source to a destination, with meaning defined by classification labels. A joint end-to-end design of the source encoder, relay transformation, and destination decoders is developed to optimize both semantic inference accuracy and perceptual reconstruction quality for image data.
By Yalin E. Sagduyu · arxiv.org
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.