Understanding the quality of calibrations for indoor localisation

Document Type

Conference Proceeding

Publication Date



© 2018 IEEE. The efficient and effective deployment of Internet of Things (IoT) systems in real world scenarios remains a challenge, particularly in applications such as indoor localisation. Various methods have been proposed recently to calibrate localisation systems, ranging from precise but time consuming processes to those involving little explicit calibration based on a crowdsourced collection of data over time. However it is not clear how to estimate and compare the quality of a specific instance of a calibration. In this paper we present a simple yet effective method of calibrating a Smart Home in a Box (SHiB) together with a framework to combine calibrations while assessing their quality. Our empirical results demonstrate that our calibration method can be performed by untrained users in a short period of time yet is capable of up to 92% accuracy in room level localisation on free living experimental data.

Publication Title

IEEE World Forum on Internet of Things, WF-IoT 2018 - Proceedings

This document is currently not available here.