Wireless Backdoor Attack and Defense for Semantic Communications over Multiple Access Channel
2026-06-29 • Networking and Internet Architecture
Networking and Internet ArchitectureCryptography and SecurityInformation TheoryMachine Learning
AI summaryⓘ
The authors study a communication system where two devices send meaningful information to one receiver at the same time. They show how an attacker can secretly send a small hidden signal to trick the receiver into misunderstanding the message from one device without affecting the other. To stop this sneaky attack, the authors create a defense that helps the receiver recognize and ignore the hidden signal to correctly interpret the messages. Their work highlights both the risk of such attacks in shared wireless communication and a possible way to guard against them.
Semantic communicationMultiple access channelLatent semantic representationBackdoor attackOver-the-air attackWireless networksSemantic inferenceTrigger-aware defenseAdversarial attackRobust training
Authors
Yalin E. Sagduyu, Tugba Erpek, Aylin Yener, Sennur Ulukus
Abstract
Semantic communication (SemCom) aims to preserve semantic meaning and task-oriented information beyond conventional message recovery over wireless channels. The adoption of SemCom in shared-access wireless networks introduces new vulnerabilities for multi-user semantic inference. This paper considers a SemCom system for two transmitters communicating with a common receiver over a multiple access channel. Each transmitter maps source information into latent semantic representations, while the receiver jointly reconstructs and classifies the semantic information for both transmitters. A selective over-the-air backdoor (Trojan) attack is presented in which an adversary transmits a low-power trigger waveform over the air and injects it into the shared received signal during training. By transmitting the trigger again during testing, this stealthy, low-power attack selectively manipulates the semantic inference for one transmitter while minimally affecting the inference of the other transmitter. To mitigate this vulnerability, a trigger-aware defense mechanism is developed to preserve correct semantic labels under trigger-contaminated wireless observations. The results demonstrate both the vulnerability of shared-access SemCom systems to selective over-the-air backdoor attacks and the effectiveness of trigger-aware robust training for semantic protection.