近日，由北京理工大学、英国萨里大学和芬兰奥卢大学组成的研究团队，提出了一种基于预训练生成式人工智能（AI）模型的低延迟、信道自适应语义通信框架。该研究成果《Latency-Aware Generative Semantic Communications with Pre-Trained Diffusion Models》已被IEEE Wireless Communication Letters（IF=4.6）录用，展示了在超低速率、低延迟和信道自适应语义通信领域的显著进展。论文第一作者为课题组博士研究生乔力。
研究背景与动机
[1] Gündüz, D., et al. "Beyond transmitting bits: Context, semantics, and task-oriented communications," IEEE J. Select. Areas Commun., vol. 41, no. 1, pp. 5--41, 2023.
[2] Chen, J., et al. "On the rate-distortion-perception function," IEEE J. Sel. Areas Inf. Theory, vol. 3, no. 4, pp. 664--673, 2022.
[3] Blau, Y., et al. "Rethinking lossy compression: The rate-distortion-perception tradeoff," Proc. Int. Conf. Mach. Learn. (ICML), pp. 675--685, 2019.
[4] Brooks, T., et al. "Video generation models as world simulators," 2024. [Online]. Available: https://openai.com/research/video-generation-models-as-world-simulators
[5] Bar-Tal, O., et al. "Lumiere: A space-time diffusion model for video generation," arXiv preprint arXiv:2401.12945, 2024.
[6] Ramesh, A., et al. "Zero-shot text-to-image generation," Proc. Int. Conf. Mach. Learn. (ICML), vol. 139, 18--24 Jul 2021, pp. 8821--8831.
所提方案概述
图1. 所提生成式语义通信框架。
01
创新点总结
02
所提方案优势
仿真结果
图2. 语义传输质量可视化结果（文本语义信息由GPT-4生成）。
生成模型对边缘图误码率的敏感性
图3. 归一化的CLIP,MS-SSIM指标与边缘图BER的关系。
信道和语义质量自适应传输
图4. 给定语义质量（BER）下，不同信噪比下最优传输参数: (a) 用于文本提示词传输的功率占比；(b) 通信传输延迟；(c) 用于边缘图传输的调制阶数（比特数、符号）；(d) 文本提示词重传的平均次数。
图5. 语义质量，调制阶数，传输/计算延迟，信噪比的对应关系表。
结论
论文信息
Essentials
标题：Latency-Aware Generative Semantic Communications with Pre-Trained Diffusion Models
作者：Li Qiao, Mahdi Boloursaz Mashhadi, Zhen Gao, Chuan Heng Foh, Pei Xiao, and Mehdi Bennis
引用格式：{L. Qiao, M. B. Mashhadi, Z. Gao, C. H. Foh, P. Xiao, and M. Bennis, ``Latency-aware generative semantic communications with pre-trained diffusion models,'' to appear in IEEE Wireless Commun. Lett. arXiv:2403.17256v1 [cs.IT], Mar 2024.}
链接：
https://arxiv.org/abs/2403.17256
谢
谢
