Scholarlink

<p>Accurate localization in Wireless Sensor Networks (WSNs) is essential to support a wide range of applications in intelligent environments. Traditional methods based on Received Signal Strength Indicator (RSSI) often encounter challenges in indoor settings due to signal propagation limitations. This study proposes a novel Distance-based Weighted Centroid-K-Nearest Neighbors (DWC-KNN) algorithm to address these issues. The algorithm combines distance-weighted centroid calculations with KNN-based inference, utilizing machine learning techniques and RSSI distance interval modeling to improve localization accuracy. Experiments conducted in various indoor environments, such as homes, libraries, and offices, demonstrate that the DWC-KNN algorithm achieves significantly lower localization errors compared to conventional geometric methods. Additionally, the transition from 2D to 3D localization, requiring four anchor nodes, marks a major step forward, enabling more precise spatial monitoring and unlocking innovative possibilities in intelligent environments. This research highlights the DWC-KNN algorithm’s effectiveness in enhancing node localization and improving the reliability and efficiency of WSN applications in complex indoor scenarios.</p>

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

Machine learning driven centroid localization algorithm for wireless sensor networks

  • 期刊论文

  • 发表日期：

    2025年7月15日

查看全文
  • Kritika Tanwar,
  • Abhinesh Kaushik

摘要

Accurate localization in Wireless Sensor Networks (WSNs) is essential to support a wide range of applications in intelligent environments. Traditional methods based on Received Signal Strength Indicator (RSSI) often encounter challenges in indoor settings due to signal propagation limitations. This study proposes a novel Distance-based Weighted Centroid-K-Nearest Neighbors (DWC-KNN) algorithm to address these issues. The algorithm combines distance-weighted centroid calculations with KNN-based inference, utilizing machine learning techniques and RSSI distance interval modeling to improve localization accuracy. Experiments conducted in various indoor environments, such as homes, libraries, and offices, demonstrate that the DWC-KNN algorithm achieves significantly lower localization errors compared to conventional geometric methods. Additionally, the transition from 2D to 3D localization, requiring four anchor nodes, marks a major step forward, enabling more precise spatial monitoring and unlocking innovative possibilities in intelligent environments. This research highlights the DWC-KNN algorithm’s effectiveness in enhancing node localization and improving the reliability and efficiency of WSN applications in complex indoor scenarios.