| 注册
首页|期刊导航|移动通信|面向工业场景语义驱动的跨域无线定位

面向工业场景语义驱动的跨域无线定位

李旭博 肖泳 李莹玉 石光明

移动通信2025,Vol.49Issue(5):67-74,8.
移动通信2025,Vol.49Issue(5):67-74,8.DOI:10.3969/j.issn.1006-1010.20250324-0002

面向工业场景语义驱动的跨域无线定位

Semantic-Driven Cross-Domain Wireless Localization for Industrial Scenarios

李旭博 1肖泳 2李莹玉 3石光明4

作者信息

  • 1. 华中科技大学电子信息与通信学院,湖北 武汉 430074
  • 2. 华中科技大学电子信息与通信学院,湖北 武汉 430074||鹏城实验室,广东 深圳 518055||琶洲实验室(黄埔),广东 广州 510335
  • 3. 中国地质大学(武汉)机械与电子信息学院,湖北 武汉 430074
  • 4. 鹏城实验室,广东 深圳 518055||琶洲实验室(黄埔),广东 广州 510335||西安电子科技大学人工智能学院,陕西 西安 710071
  • 折叠

摘要

Abstract

Contactless device-free wireless localization technology has recently attracted significant interest due to its potential to support accurate localization by using ubiquitous radio frequency signals.However,traditional solutions focus on developing a single and fixed global model,which fails to realize accurate localization under dynamic and complex industrial environment.It is also unrealistic to construct individual models for all the possible locations and scenarios due to the barriers to obtaining labeled data and cost bottlenecks.Motivated by the observation that signals recorded at different locations are closely related to environmental semantic features,this paper proposes a cross-domain wireless localization framework based on semantic-aware networks(SAN),which characterizes the intrinsic mapping between the environmental semantic features and location-aware models cross different industrial scenarios.A SAN-based zero-shot transfer solution is introduced to allow receivers in novel industrial scenarios to obtain location-specific models by directly aggregating the well-trained models in historical industrial scenarios without any labeled data.Experimental results demonstrate that the proposed SAN-based localization framework can achieve effective cross-domain localization under zero-shot conditions,thereby reducing the training and deployment costs of location-aware models.

关键词

语义认知网络/环境语义/工业互联网/无线室内定位/无监督域适应

Key words

semantic-aware network/environment semantics/industrial internet/wireless indoor localization/unsupervised domain adaptation

分类

信息技术与安全科学

引用本文复制引用

李旭博,肖泳,李莹玉,石光明..面向工业场景语义驱动的跨域无线定位[J].移动通信,2025,49(5):67-74,8.

基金项目

移动信息网络国家科技重大专项项目"面向语义通信的语义知识库技术研究"(2024ZD1300700) (2024ZD1300700)

国家自然科学基金"模型驱动的异构联邦边缘智能网络碳排放建模与优化"(62301516) (62301516)

湖北省国际科技合作计划项目"面向印尼通信工程复杂场景的云边协同智能运维系统研发及应用"(2023EHA009) (2023EHA009)

中国地质大学(武汉)"地大学者"人才岗位科研启动经费资助(2021164) (武汉)

移动通信

1006-1010

访问量6
|
下载量0
段落导航相关论文