Tomoaki Ogawa

A new in-door location detection method adopting learning algorithms

This new location detection method uses wireless LAN systems. The location of the user station is established using the received intensity at a user station of signals from wireless base stations. The method employs two unique concepts. 1) Instead of a discrete coordinate position to define the location of user station, the location is found by using a localized area. 2) The learning algorithms based on LVQ (learning vector quantization) are applied to process in location. As a result, our method provides more accurate positions compared with conventional least squares method (LSM) methods in areas far from the base points.

Picocell network for local positioning and information system

Geographic information services are becoming attractive with the development of mobile computing. For these services, it is necessary to provide highly accurate positioning detection, but existing systems fail to satisfy all requirements. This paper introduces a highly accurate positioning information system that is realized by small transmitters. We show some examples of geographic information services and the functions required. A system architecture that satisfies the requirements and technical subjects is then presented. The outline of a recent prototype is described.