arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2501.04732cs.ITcs.AImath.IT
收藏

SNR-EQ-JSCC:基于SNR嵌入与查询的联合信源信道编码

SNR-EQ-JSCC: Joint Source-Channel Coding with SNR-Based Embedding and Query

Shanghai Jiao Tong University(上海交通大学)

Hongwei Zhang, Meixia Tao

AI总结:

针对JSCC语义通信中的动态信道适配问题,提出基于SNR嵌入和信道自适应查询的轻量Transformer架构SNR-EQ-JSCC,以极低开销提升图像传输性能。

摘要:

应对动态信道的影响是基于联合信源信道编码(JSCC)的语义通信系统中的一个关键问题。本文提出一种轻量级信道自适应语义编码架构,称为SNR-EQ-JSCC。该架构构建于通用Transformer模型之上,通过将信噪比(SNR)嵌入注意力块,并借助信道自适应查询动态调整注意力分数,从而实现信道自适应(CA)。同时,在损失函数中引入惩罚项以稳定训练过程。考虑到瞬时SNR反馈可能不完美,本文提出一种仅使用平均SNR的替代方法,且无需重新训练SNR-EQ-JSCC。在图像传输上开展的仿真结果表明,所提出的SNR-EQ-JSCC在峰值信噪比(PSNR)和感知指标上优于最先进的SwinJSCC,而实现CA仅需0.05%的存储开销和6.38%的计算复杂度。此外,信道自适应查询方法在感知指标上展现出显著提升。当瞬时SNR反馈不完美时,仅使用平均SNR的SNR-EQ-JSCC仍优于基线方案。

英文摘要:

Coping with the impact of dynamic channels is a critical issue in joint source-channel coding (JSCC)-based semantic communication systems. In this paper, we propose a lightweight channel-adaptive semantic coding architecture called SNR-EQ-JSCC. It is built upon the generic Transformer model and achieves channel adaptation (CA) by Embedding the signal-to-noise ratio (SNR) into the attention blocks and dynamically adjusting attention scores through channel-adaptive Queries. Meanwhile, penalty terms are introduced in the loss function to stabilize the training process. Considering that instantaneous SNR feedback may be imperfect, we propose an alternative method that uses only the average SNR, which requires no retraining of SNR-EQ-JSCC. Simulation results conducted on image transmission demonstrate that the proposed SNR-EQJSCC outperforms the state-of-the-art SwinJSCC in peak signal-to-noise ratio (PSNR) and perception metrics while only requiring 0.05% of the storage overhead and 6.38% of the computational complexity for CA. Moreover, the channel-adaptive query method demonstrates significant improvements in perception metrics. When instantaneous SNR feedback is imperfect, SNR-EQ-JSCC using only the average SNR still surpasses baseline schemes.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑