A multilayered space-event model for navigation in indoor spaces

In this paper a new conceptual framework for indoor navigation is proposed. While route planning requires models which reflect the internal structure of a building, localization techniques require complementary models reflecting the characteristics of sensors and transmitters. Since the partitioning of building space differs in both cases, a conceptual separation of different space models into a multilayer representation is proposed. Concrete space models for topographic space and sensor space are introduced. Both are systematically subdivided into primal and dual space on the one hand and (Euclidean) geometry and topology on the other hand. While topographic space describes 3D models of buildings and their semantically subdivisions into storey's and rooms, sensor space describes the positions and ranges of transmitters and sensors like Wi-Fi access points or RFID sensors. It is shown how the connection of the different layers of the space models describe a joint state of a moving subject or object and reduces uncertainty about its current position.