Semantic Feature Division Multiple Access for Digital Semantic Broadcast Channels

TL;DR

This paper introduces a novel semantic feature division multiple access (SFDMA) scheme for multi-user broadcast channels, enabling simultaneous transmission of semantic information with privacy protection and optimized performance for inference and image reconstruction tasks.

Contribution

The paper proposes a new SFDMA paradigm that encodes multi-user semantic info into orthogonal representations, improving multi-user interference management and privacy in broadcast channels.

Findings

SFDMA achieves a tradeoff between inference performance, data compression, and interference.

The use of Swin Transformer significantly reduces multi-user interference in image reconstruction.

Simulations confirm the effectiveness and superiority of the proposed SFDMA scheme.

Abstract

In this paper, we propose a digital semantic feature division multiple access (SFDMA) paradigm in multi-user broadcast (BC) networks for the inference and the image reconstruction tasks. In this SFDMA scheme, the multi-user semantic information is encoded into discrete approximately orthogonal representations, and the encoded semantic features of multiple users can be simultaneously transmitted in the same time-frequency resource. Specifically, for inference tasks, we design a SFDMA digital BC network based on robust information bottleneck (RIB), which can achieve a tradeoff between inference performance, data compression and multi-user interference. Moreover, for image reconstruction tasks, we develop a SFDMA digital BC network by utilizing a Swin Transformer, which significantly reduces multi-user interference. More importantly, SFDMA can protect the privacy of users' semantic information, in which each receiver can only decode its own semantic information. Furthermore, we establish a relationship between performance and signal to interference plus noise ratio (SINR), which is fitted by an Alpha-Beta-Gamma (ABG) function. Furthermore, an optimal power allocation method is developed for the inference and reconstruction tasks. Extensive simulations verify the effectiveness and superiority of our proposed SFDMA scheme.