The effects of dynamic environment on channel frequency response-based indoor positioning

Abstract

Indoor Positioning Systems should be able to locate an object or person in a dynamic environment. In this research work, we present experimental results when performing indoor positioning in a dynamic environment when using channel frequency response as the indoor location's fingerprint. Dynamic environments introduce varying channel frequency responses for a given indoor location, which makes it difficult to locate desired indoor objects. According to our experiments, we find that indoor fingerprints become uncorrelated within a 15-min window. To bridge the gap between daily fingerprints, we propose three quick remedies in updating the Channel Frequency Response (CFR) database. These updates allow indoor localization even in the presence of dynamic environmental changes (e.g., people movement, interference of other wireless signals, etc.) and are able to 100% localize indoor locations separated by at least five (5) centimeters. We highlight that in the experiments, there is only one anchor node used to estimate the desired indoor location.