Utility based node selection scheme for cooperative localization

Cooperative positioning is a popular approach for indoor localization, which significantly outperforms conventional positioning techniques. This paper presents a method for iterative node localization, with a relatively small number of initial anchor nodes. In order to reduce error propagation in iterative schemes, it is necessary to use only reliable information among nodes. We introduce a reference node selection strategy based on utility functions. The method is completely distributed and involves only information from local neighborhood. Compared to the nearest neighbor and random node selection scheme, simulation results show that selecting subsets of nodes with highest utility values leads to more accurate position estimates.