Machine Learning and its Applications in Visible Light Communication Based Indoor Positioning

Abstract

Visible light communication (VLC) based indoor positioning technique has shown significant advantages including low cost, high durability and good environmental friendliness. In this paper, we present a comprehensive perspective on current VLC based indoor positioning systems and analyze existing obstacles to realize high precision. Then some promising machine learning techniques utilized in indoor positioning are discussed including KNN, ANN, clustering and fusion of multiple classifiers, with their performance in accuracy, robustness and efficiency also analyzed and compared in detail. Two future typical application scenes are further proposed at last.