Wi-Fi Fingerprint localisation using Density-based Clustering for public spaces: A case study in a shopping mall

Abstract

Indoor localisation is to-date still an active research area. This paper presents a case study on a localisation technique using Wi-Fi fingerprints built from radio information collected using commercially-off-the-shelf smartphones. The Wi-Fi fingerprints are built using density-based clustering-based algorithms. The investigation is carried out on normal operation scenarios, where a normal crowd was present during the experiments. A simplified version of the clustering algorithm, the Simplified Fingerprint Density-based Clustering Algorithm (SFDCA), is proposed, implemented as well as evaluated with a comparison to an existing indoor localisation algorithm called Density-based Cluster Combined Algorithm (DCCLA). Furthermore, a few changes are proposed and evaluated for the recognition algorithm. This paper discusses the obtained results, observations and issues faced in the case study.