A Characteristics-Based Least Common Multiple Algorithm to Optimize Magnetic-Field-Based Indoor Localization

• Published in: IEEE internet of things journal Vol. 12; no. 8; pp. 10463 - 10478
• Main Authors: Rafique, Hamaad, Patti, Davide, Palesi, Maurizio, La Delfa, Gaetano Carmelo
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 15.04.2025; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Accuracy; Algorithms; Automatic clustering; Automation; Classification; Clustering; Clustering algorithms; Clustering methods; data clustering; Fingerprint recognition; indoor localization; Internet of Things; least common multiple (LCM); Localization; Location awareness; Machine learning; novel clustering technique; Real time; Real-time systems; Shape; Time series analysis; Unsupervised learning; Wireless fidelity
• ISSN: 2327-4662; 2327-4662
• DOI: 10.1109/JIOT.2024.3511261

Clustering is an unsupervised learning technique that groups data based on similarity criteria. Traditional methods like K-Means and agglomerative clustering often require predefined parameters, struggle with irregular cluster shapes, and fail to classify subcluster points in magnetic fingerprint-based indoor localization. This study proposes the characteristics-based least common multiple (LCM) algorithm to address these challenges. This novel approach autonomously determines cluster number and shape while accurately classifying misclassified points based on characteristic similarities using LCM. We evaluated the proposed technique using state-of-the-art metrics and tested it in magnetic-field-based indoor localization scenarios. Comparisons were made with real-time and benchmark datasets, alongside traditional clustering methods. Results demonstrate that LCM significantly enhances localization accuracy, achieving a mean absolute error rate of 0.1 m.