Indoor positioning application development on an Android smart-phone

Abstract

Location-aware services have become popular with the development of modern communication technology. The increased variety of commercial applications has established the demand for indoor localization services.

Several indoor localization systems have been developed that rely on infrared, ultra sound or ultra wide band technologies inside a building. Although these technologies deliver high positioning accuracy, they require the deployment of expensive hardware. Today, most indoor positioning systems rely on wireless local area networks (WLAN) owing to the widely available existing infrastructure, i.e., WLAN Access Points (AP), and coverage inside buildings.

Moreover, commercial mobile devices can be used without any hardware modifications as they are usually equipped for wireless connectivity. Many WLAN based indoor positioning systems have been developed using measurement techniques like angle of arrival, angle difference of arrival, time of arrival and time difference of arrival. However, these techniques are suitable for systems with direct line-of-sight and have problems or require complex computation in indoor environments due to noise, interference, and multipath. Hence, the fingerprinting technique has gained attention due to its simplicity as compared to other techniques.

In this project a fingerprinting based indoor positioning system using existing infrastructure is developed. The system attempts to address the drawbacks of current fingerprinting based positioning systems. It enables user privacy with low communication overhead by allowing the user‘s mobile device to determine its own position. It also enables efficient update of the fingerprint database through ‘Crowd-Sourcing’. In addition, it allows users to create location based applications specifying conditions under which specific actions can be triggered. The developed system however has highly inaccurate position estimates. This is because the system faces several challenges like severe fluctuation in the strength of Wi-Fi signals and the lack of standardization of the devices carried by the consumers of location services. Further optimizations like hyperbolic location fingerprinting or dead reckoning can be added to the system to overcome these challenges and increase accuracy to a suitable level.