Mobile and redundant access point reduction for indoor unmanned aerial vehicle positioning using WLAN crowdsourcing fingerprints

Abstract

In recent years, due to the rapid development of indoor Unmanned Aerial Vehicles (UAVs), the positioning of indoor UAVs has played a significant role in UAVs applications. Among the existing approaches, the Wireless Local Area Network (WLAN) based positioning approach is recognized as an effective way by the benefit from the widely-deployed WLAN. At the same time, the continued expansion of WLAN is accompanied by the numerous types of Access Points (APs) such as mobile phones and tablets. In this circumstance, the existence of mobile APs reduces the location dependency of Received Signal Strength (RSS) data, and meanwhile the large number of APs results in huge storage and computational overhead to the positioning. In response to these compelling problems, we propose a joint judgment criterion by which the WLAN crowdsourcing fingerprints are applied to detect mobile APs. After that, we use the neighborhood rough set reduction algorithm to calculate the weight of each AP and then remove the APs with zero weight. Finally, we rely on the fingerprints from remaining APs to construct a fingerprint database for the positioning. The extensive experimental results show that the proposed approach can precisely detect mobile APs as well as guarantee the satisfactory positioning accuracy under the redundant APs removal.