Mobile sensor mapping via semi-definite programming

We consider the problem of mapping the locations of a mobile device into the Euclidean space utilizing its perception of the environment through sensors, e.g., Wi-Fi. We formulate the estimation problem as a less-effort dynamic fringerprint technique, which capitalizes to non-convex optimization problem. Specifically, we leverage spatial correlation models and properties of the Euclidean Distance Matrices to derive a Semi Definite Programming (SDP) formulation of a GP-LVM-like estimation problem. The proposed algorithm has been tested with real WiFi measurements in indoors and results show good performance in capturing the shape and size of the walked trajectory.