A Location-Aware User Tracking and Prediction System Abstract— Modern location-aware services and applications use context and prediction methods to adapt to the needs of users and changes in the environment. The growing availability of WLAN and mobile devices offers significant opportunities for location- aware services. But the use of WLAN or RFID technologies alone provides a less accurate estimation of a user’s current location. In this paper, we introduce an ontology-based location tracking system. It makes use of both WLAN and RFID technologies, and includes a prediction method for identifying a user’s current location and predicted future location. Our system architecture better serves the client by using location and context information. Keywords- context-aware; location-aware; ontology; Shafer- theory; ubiquitous environment. I. INTRODUCTION Location-awareness plays a major role in enhancing ubiquitous environments. It delivers services to mobile users based on personal and general context according to the location they are in. Many applications developed today, such as medical services, home applications, office applications, and university services, are enhanced by the use of location and context information. The most common method of tracking location is the Global Positioning System (GPS). The drawback is that GPS cannot be used indoors. Other methods must therefore be considered. A combination of IEEE 802.11 WLAN and RFID tagging technologies allows mobile users to be tracked and services to be deployed both indoors and outdoors. The goal of our system is to enable each user to interact with the surroundings based on his or her location. Systems of this kind can be used for ontology-based negotiations [1], seamless video handoff [2], and mobility prediction [3]. They can also display a map of the surroundings to guide the user inside a building, or print a document on the nearest printer. The interaction varies as the location changes. For example, when the user is in his or her office, the mobile device may receive work data only, but, if the user moves to the cafeteria, data may include information on friends and entertainment. This requires the consideration of context, described as characteristics that are physically and logically measurable, such as location, objects, services, applications, and the like. Our research explores two specific technologies for location sensing; WLAN and RFID. The first approach consists of discovering the surrounding access points and measuring the signal strength sent to mobile devices by each access point. The second approach improves accuracy by using active RFID tagging, where signal strength sent from the RFID reader are measured. In both cases, the signal strength is then transferred to a two-dimensional map. Once a user’s current location has been identified, a “prediction center” identifies the user’s next location. The “next location prediction” method is used so that the system can take actions before the user arrives at the destination. Those actions involve both the network level (such as network handover management [4]), and the service level (such as ontology-based negotiations and seamless video handoff). The “Dempster-Shafer Theory” [5] is used for prediction because of its ability to gather pieces of evidence that help it select from different possible future locations. In this paper, we propose an architecture for location- awareness by integrating two location technologies. It handles information from both WLAN access points and RFID readers and locates users from those readings. The users’ future locations are estimated using the Dempster-Shafer Theory. The rest of this paper is organized as follows: In section 2, we review some of the related work in the field and its limitations. In section 3, we present and discuss our system architecture. In section 4, we apply the Dempster-Shafer Theory to our system in order to predict future locations. Section 5 illustrates the use of the system in a campus setting. Section 6 briefly explains our implementation prototype. Section 7 discusses the evaluation and simulation results. Section 8 concludes the paper with ideas for future work. II. RELATED WORK Location-awareness in mobile environments is an important research area with a number of approaches. Work in the area makes use of different technologies to locate and track an object, including infrared, radio frequency, ultrasound, magnetic fields, and cellular systems. Rao et al. [6] describe platforms, technologies and standards used for location-based services (LBS). Technologies such as PDAs need to be connected and integrated with other infrastructures such as wireless and satellite networks to achieve success. Knowledge of the user’s location is only one I. Al Ridhawi 1, M. Aloqaily 1 , A. Karmouch 1 , N Agoulmine 2 1 School of Information Technology and Engineering (SITE), University of Ottawa, PO Box 450, Ottawa, ON, K1N 6N5, Canada. 2 LRSM – ENSIIE, 18 Allée Jean Rostand, 91025, Evry, France Ialri083@uottawa.ca , karmouch@site.uottawa.ca 978-1-4244-4624-7/09/$25.00 ©2009 IEEE