OPEN Wireless localization with diffusion maps

This paper proposes a novel approach to the Wireless Localization Matching Problem (WLMP) using diffusion maps, a manifold learning technique. The WLMP involves matching sensor nodes with known positions based on noisy wireless signal measurements. The authors demonstrate that diffusion maps effectively embed positions and equipment coordinates in a space enabling reliable coordinate comparison and assignment quality evaluation with low computational cost. The method is shown to be robust to noise, achieving accurate matching even in realistic scenarios with low signal-to-noise ratios. This suggests significant potential for increasing the accuracy of wireless localization.

Amin Ghafourian, Orestis Georgiou, Edmund Barter, Thilo Gross