arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2606.01899eess.SPcs.AI
收藏

RA-LWLM:基于检索增强的上下文无线定位基础模型

RA-LWLM: Retrieval-Augmented In-Context Localization with Wireless Foundation Models

Department of Electrical Engineering, Chalmers University of Technology(查尔姆斯理工大学电子工程系);Department of Electrical and Computer Engineering, Aarhus University(阿鲁斯大学电子与计算机工程系)

Guangjin Pan, Hui Chen, Hei Victor Cheng, Henk Wymeersch

AI总结:

提出RA-LWLM框架,通过将场景特定信息外化到指纹数据库,实现无需训练的跨场景无线定位,利用冻结的无线基础模型编码器、检索模块和基于Transformer的上下文学习模块预测用户位置。

摘要:

无线定位是第六代(6G)网络的基本能力。传统的基于模型的方法需要对传播环境进行精确建模,在复杂的多径和非视距场景中性能下降,而基于学习的方法将模型参数紧密耦合到训练场景中,每当基站(BS)配置或传播环境变化时需要昂贵的重新训练。在本文中,我们提出RA-LWLM,一种检索增强的上下文定位框架,通过将场景特定信息外化到每个场景的指纹数据库(而非编码在模型权重中)来实现无需训练的跨场景适应。该框架由三个组件组成:一个冻结的无线基础模型(FM)编码器,将原始信道状态信息映射为场景无关的表示;一个检索模块,通过表示空间中的相似性搜索从每个场景的数据库中选择最具信息量的参考;以及一个基于Transformer的上下文学习(ICL)模块,将查询与检索到的参考融合以预测用户设备(UE)位置。为了适应不同查询的检索质量和传播复杂性,ICL模块采用混合专家设计,其中专家专注于不同的上下文大小,并由可学习的选择器软组合。跨不同BS配置的异构场景的广泛基于射线追踪的实验表明,RA-LWLM在未见和已见场景上实现了几乎相同的精度,无需任何每个场景的重新训练,显著优于端到端和基于FM的基线。这些结果验证了所提出的检索增强上下文范式作为6G网络中跨场景定位的可扩展解决方案。

英文摘要:

Wireless localization is a fundamental capability of sixth-generation (6G) networks. Conventional model-based methods require accurate modeling of the propagation environment and degrade in complex multipath and non-line-of-sight scenarios, while learning-based methods couple model parameters tightly to the training scene, requiring costly retraining whenever the base station (BS) configuration or propagation environment changes. In this paper, we propose RA-LWLM, a retrieval-augmented in-context localization framework that achieves training-free cross-scene adaptation by externalizing scene-specific information into a per-scene fingerprint database rather than encoding it in model weights. The framework consists of three components: a frozen wireless foundation model (FM) encoder that maps raw channel state information into a scene-agnostic representation; a retrieval module that selects the most informative references from the per-scene database via similarity search in the representation space; and a transformer-based in-context learning (ICL) module that fuses the query with the retrieved references to predict the user equipment (UE) position. To accommodate varying retrieval quality and propagation complexity across queries, the ICL module adopts a mixture-of-experts design in which experts specialize in different context sizes and are softly combined by a learnable selector. Extensive ray-tracing-based experiments across heterogeneous scenes with diverse BS configurations show that RA-LWLM achieves nearly identical accuracy on seen and unseen scenes without any per-scene retraining, substantially outperforming end-to-end and FM-based baselines. These results validate the proposed retrieval-augmented in-context paradigm as a scalable solution for cross-scene localization in 6G networks.

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。

↑