Secure Semantic Communication via Paired Adversarial Residual Networks

TL;DR

This paper proposes a novel semantic communication system that uses paired adversarial residual networks to enhance security by fooling eavesdroppers and maintaining high communication quality.

Contribution

It introduces a joint optimization framework for trainable ARNs at transmitter and receiver to defend against eavesdropping while ensuring semantic communication performance.

Findings

Successfully fools eavesdroppers in simulations

Maintains high semantic communication quality

Effectively balances attack power and communication accuracy

Abstract

This letter explores the positive side of the adversarial attack for the security-aware semantic communication system. Specifically, a pair of matching pluggable modules is installed: one after the semantic transmitter and the other before the semantic receiver. The module at transmitter uses a trainable adversarial residual network (ARN) to generate adversarial examples, while the module at receiver employs another trainable ARN to remove the adversarial attacks and the channel noise. To mitigate the threat of semantic eavesdropping, the trainable ARNs are jointly optimized to minimize the weighted sum of the power of adversarial attack, the mean squared error of semantic communication, and the confidence of eavesdropper correctly retrieving private information. Numerical results show that the proposed scheme is capable of fooling the eavesdropper while maintaining the high-quality semantic communication.