Virtual and Oriented WiFi Fingerprinting Indoor Positioning based on Multi-Wall Multi-Floor Propagation Models

摘要

Virtual fingerprints have been proposed in the context of WiFi Fingerprinting Indoor Positioning systems in order to reduce the effort dedicated to offline measurements. In this work, the use of Multi-Wall Multi-Floor indoor propagation models to generate such virtual fingerprints is investigated. A strategy taking into account the impact of user/device orientation on the signal propagation is proposed, leading to the creation of virtual and oriented fingerprints. The work analyzes then the trade-offs between model accuracy and measurement efforts by means of experimental results, showing that good modeling accuracy can be guaranteed while significantly reducing the complexity of the offline measurement phase.