OPEN Wireless localization with diffusion maps

In the Wireless Localization Matching Problem (WLMP) the challenge is to match pieces of equipment with a set of candidate locations based on wireless signal measurements taken by the equipment. This is complicated by noise inherent in wireless signal measurements. The paper proposes using diffusion maps, a manifold learning technique, to embed positions and equipment coordinates into a space that enables coordinate comparison and reliable evaluation of assignment quality at very low computational cost. The mapping is shown to be robust to noise and enables accurate matching in realistic settings, suggesting that diffusion-map-based approaches can significantly increase the accuracy of wireless localization in applications.