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 2511.14895eess.SP
收藏

面向边缘设备无线时间序列下游任务的轻量级基础模型

Lightweight Foundation Model for Wireless Time Series Downstream Tasks on Edge Devices

Ghent University(根特大学);Yonsei University(延世大学);Khalifa University of Science and Technology(哈利法科学与技术大学)

Mohammad Cheraghinia, Eli De Poorter, Jaron Fontaine, Kwang Soon Kim, Merouane Debbah, Adnan Shahid

AI总结:

针对现有无线领域基础模型计算开销大难以部署在边缘设备的问题,本文提出基于MLP编码器的轻量级无线时间序列基础模型,仅21K参数即可支持4类下游任务,泛化性能稳健且推理速度快。

摘要:

尽管机器学习已被广泛用于优化无线网络,但由于训练与部署成本高昂,为通信和定位领域的每项任务单独训练模型的模式正变得难以持续。基础模型可通过少量样本微调让单个模型适配多类任务,提供了更具可扩展性的替代方案。然而现有基础模型大多依赖大规模Transformer架构,计算开销极高,不适用于典型边缘设备部署。本文提出一种基于简单多层感知机(MLP)编码器的轻量级基础模型,各编码器可独立处理输入补丁。该模型支持来自IQ、CIR等多种输入类型、不同采样率下的4类下游任务:长距离技术识别、短距离技术识别、调制识别、视距检测。实验表明,与随下游任务增加会出现性能下降的Transformer不同,该MLP模型可保持稳健的泛化性能,对先前未见过的数据类别微调准确率超过97%。上述结果是在模型仅含2.1万个可训练参数的条件下实现的,其在普通边缘设备上的推理时间仅为0.33毫秒,适用于资源受限的实时部署场景。

英文摘要:

While machine learning is widely used to optimize wireless networks, training a separate model for each task in communication and localization is becoming increasingly unsustainable due to the significant costs associated with training and deployment. Foundation models offer a more scalable alternative by enabling a single model to be adapted across multiple tasks through fine-tuning with limited samples. However, current foundation models mostly rely on large-scale Transformer architectures, resulting in computationally intensive models unsuitable for deployment on typical edge devices. This paper presents a lightweight foundation model based on simple Multi-Layer-Perceptron (MLP) encoders that independently process input patches. Our model supports 4 types of downstream tasks (long-range technology recognition, short-range technology recognition, modulation recognition and line-of-sight-detection) from multiple input types (IQ and CIR) and different sampling rates. We show that, unlike Transformers, which can exhibit performance drops as downstream tasks are added, our MLP model maintains robust generalization performance, achieving over 97% accurate fine-tuning results for previously unseen data classes. These results are achieved despite having only 21K trainable parameters, allowing an inference time of 0.33 ms on common edge devices, making the model suitable for constrained real-time deployments.

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。

↑