Enhanced GNSS Positioning Solution on Android for Location Based Services Using Big Data

Abstract

GNSS based positioning is common and widely utilized positioning system in world. The only limitation beside of accuracy a GNSS receiver has is the availability of GNSS signal. Closed environments where direct view of sky is not available e.g., inside building or shopping mall or under tunnel do not offer valid GNSS positioning. These types of environments are generally termed as GNSS denied environments (GDEs). In such scenario a modified technology called Dead Reckoning System (DRS) is used for positioning. A DRS is a sophisticated and costly system which is not widely available for general public usage. In this paper, we propose a novel and state of the art DRS technology. Specifically, we develop the DRS on Android platform for providing positioning services in real time using state of the art, open-source big data technologies. To ensure effective and accurate positioning with minimum power and resource consumption. We use the in-memory computing and streaming of Apache Spark and data storage of Aerospike. We test our DRS in three different GDEs along with the PoC in which we compare our DRS result with actual GPS-based positions. In each experiment, our system has given a promising circular error probability (CEP) of less than 3 meters, irrespective of object state. The potential domains of applications are security, location-based service, and indoor navigation.