Lightitude: Indoor Positioning Using Uneven Light Intensity Distribution.

In this paper, we propose an indoor positioning system, Lightitude, which utilizes the already existed, uneven indoor light intensity distribution established by densely deployed indoor lights as the medium. As common indoor lights cannot act as landmarks due to lack of unique features (e.g., unique intensity or flicker frequency), systems that exploiting the received light intensity (RLI) usually impose strong constraints on user's motion and make ideal assumptions about the indoor environment. Different from these approaches, we first propose a realistic light intensity model to reconstruct the RLI distribution given any motion (position, orientation) of the receiver, thus RLI collected with every motion of the receiver could be used for positioning. Then we design a particle-filter-based positioning module, which harnesses user's natural mobility to eliminate the ambiguity of a single RLI. Experiment results show that Lightitude achieves mean accuracy 1.93m and 1.98m in an office (720m2) and a library (960m2) respectively. Lightitude is still robust against interferences like sunlight, shading of human-body and several user behaviors.