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 2310.09394cs.LGcs.AIcs.ITcs.NImath.IT
收藏

基于分割学习的语义对齐用于弹性多用户语义通信

Semantics Alignment via Split Learning for Resilient Multi-User Semantic Communication

Yonsei University(延世大学);Deakin University(迪肯大学);Inha University(仁荷大学);University of Oulu(奥卢大学)

Jinhyuk Choi, Jihong Park, Seung-Woo Ko, Jinho Choi, Mehdi Bennis, Seong-Lyun Kim

AI总结:

针对多用户语义通信中神经收发器因数据与信道差异导致的语义不对齐问题,提出基于分割学习与层冻结微调的SLF方法,通过调整微调层比例控制成本,仿真验证其能有效对齐语义。

摘要:

近期关于语义通信的研究通常依赖于基于神经网络(NN)的收发器,例如深度联合源信道编码(DeepJSCC)。与传统收发器不同,这些神经收发器可利用实际源数据和信道进行训练,使其能够提取和传达语义。另一方面,每个神经收发器固有地偏向于特定的源数据和信道,使得不同收发器难以理解预期的语义,尤其是在初次相遇时。为了在多个神经收发器之间对齐语义,我们提出了一种基于分布式学习的解决方案,该方案利用分割学习(SL)和部分神经网络微调技术。在这种被称为层冻结分割学习(SLF)的方法中,每个编码器下载一个未对齐的解码器,并在本地微调这些编码器-解码器神经网络层的一部分。通过调整这一部分,SLF可以控制计算和通信成本。仿真结果证实了SLF在不同源数据和信道差异下对齐语义的有效性,体现在分类准确率、重建误差以及从未对齐状态理解预期语义的恢复时间方面。

英文摘要:

Recent studies on semantic communication commonly rely on neural network (NN) based transceivers such as deep joint source and channel coding (DeepJSCC). Unlike traditional transceivers, these neural transceivers are trainable using actual source data and channels, enabling them to extract and communicate semantics. On the flip side, each neural transceiver is inherently biased towards specific source data and channels, making different transceivers difficult to understand intended semantics, particularly upon their initial encounter. To align semantics over multiple neural transceivers, we propose a distributed learning based solution, which leverages split learning (SL) and partial NN fine-tuning techniques. In this method, referred to as SL with layer freezing (SLF), each encoder downloads a misaligned decoder, and locally fine-tunes a fraction of these encoder-decoder NN layers. By adjusting this fraction, SLF controls computing and communication costs. Simulation results confirm the effectiveness of SLF in aligning semantics under different source data and channel dissimilarities, in terms of classification accuracy, reconstruction errors, and recovery time for comprehending intended semantics from misalignment.

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。

↑