Wireless Personal Communications 30: 217–231, 2004. C  2004 Kluwer Academic Publishers. Printed in the Netherlands. An Integrated Approach for the Estimation of Mobile Subscriber Geolocation PANAGHIOTIS KIKIRAS 1 and DIMITRIS DRAKOULIS 2 1 Senior Researcher, Department of Electrical and Computer Engineering, National Technical University of Athens, Zwgrafou 15773 E-mail: pkikiras@army.gr, kikirasp@ieee.org 2 Senior Researcher, Telecommunications Laboratory, National Technical University of Athens, Zwgrafou 15773 E-mail: ddrakoul@telecom.ntua.gr Abstract. The continuous growth of wireless services market, fuels the need for precise location dependent services, leading researchers from academia and industry to reassess existing geolocation methods regarding accuracy and availability of position estimation. The proposed method for mobile subscriber geolocation utilizes key concepts from estimation theory and specifically the Kalman filter algorithm to determine an optimal estimate on the actual system state (which primarily includes location, velocity) based on the observations acquired by employing network- or terminal-based techniques, which are briefly presented and assessed thereafter. Given the proven limitations of individual techniques, the alternative strategies for fusion of data are outlined, the details of the operation of a fusion scheme based on the Kalman filter are discussed and the impact of the proposed work over conventional methodologies is quantified. Keywords: mobile geolocation, state estimation, Kalman filtering, data fusion 1. Introduction Operators and service providers alike, continuously seek innovative ways to create service differentiation and added revenue opportunities through the delivery of highly personalized services. On the other hand, subscribers increasingly demand for added value. The determina- tion of a subscriber’s location – also referred to as “geo-location” – is one of the most powerful ways to personalize services and provide added value for the end-user. Locating a mobile ter- minal (MT) may be extremely useful in cases of emergency, traffic and navigation assistance are seen as great assets to drivers, tourists and travelers, while multiple other information services may also be offered. In the US case, conformance to Federal Communications Com- mission’s (FCC) “E911” telecommunication initiatives requires that wireless phone providers develop methods to accurately locate any phone which places a 911 emergency call [1]. Precise mobile callers’ geo-location to within 150 m for 95% of calls is required [2]. Similar to the FCC, in Europe the Coordination Group on Access to Location Information for Emergency Services (CGALIES) [3] has set forward requirements which were to be finalized within 2003. The requirements are not dependent on the emergency service but on the environment (rural, urban) and on the situation (visibility, ability of the user to provide information). However, pre- cise (according to FCC standards) and uniform geolocation information provision has proven a challenging task and results show that no individual method suffices for uniform results. Methodologies and algorithms for fusion of multiple sensor data are thus required to improve precision as well as availability and fault tolerance.