A graph-based approach for mobile localization exploiting real and virtual landmarks

In the last decade, several localization systems have been proposed, some of them achieving an accuracy in the order of decimeters. Nonetheless, a broad range of these methods may reveal unattractive for low-cost and resource-bounded mobile application scenarios. Indeed, significant constraints stem from costs and availability of the required technological infrastructure for the target environment. In addition, resources necessary to train and run the localization algorithm may pose stringent requirements in terms of time, power and computational. In the aforementioned context, this paper presents an effective radio–based positioning algorithm able to localize the user by exploiting a graph-based representation of the environment. The graph is created by analyzing the received signal level and the movement directions of a mobile device with respect to a set of real and virtual landmarks. The comparison with a commercial localization system proves the effectiveness and accuracy of the proposed approach.