Scholarlink

Utilizing Wi-Fi signals for indoor localization significantly improves location-based services in indoor environments, though challenges arise due to unpredictable Wi-Fi signal propagation. We propose an innovative Feature Enhancement and K-Nearest Neighbor (FEKNN) approach, which refines RSSI data distribution for a more accurate feature database and employs a refined Weighted K-Nearest Neighbor (W-KNN) algorithm to calculate locations by Euclidean distances between enhanced features. Extensive experiments validate that our FEKNN has remarkable accuracy for indoor localization applications, achieving state-of-the-art performance with an impressive average localization error of 1.86 meters on the public UjiIndoorLoc testing dataset, and an average error of 0.68 meters on our custom-built dataset.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

FEKNN: A Wi-Fi Indoor Localization Method Based on Feature Enhancement and KNN

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Jingqi Wang,
  • Jinming Yang,
  • Bowen Li,
  • Weiliang Meng,
  • Jiguang Zhang,
  • Xiaopeng Zhang

摘要

Utilizing Wi-Fi signals for indoor localization significantly improves location-based services in indoor environments, though challenges arise due to unpredictable Wi-Fi signal propagation. We propose an innovative Feature Enhancement and K-Nearest Neighbor (FEKNN) approach, which refines RSSI data distribution for a more accurate feature database and employs a refined Weighted K-Nearest Neighbor (W-KNN) algorithm to calculate locations by Euclidean distances between enhanced features. Extensive experiments validate that our FEKNN has remarkable accuracy for indoor localization applications, achieving state-of-the-art performance with an impressive average localization error of 1.86 meters on the public UjiIndoorLoc testing dataset, and an average error of 0.68 meters on our custom-built dataset.