Sparse Point Source Estimation in Sensor Networks with Gaussian Kernels

In this paper we present a technique for the estimation of point sources in a diffusive environment that can be applied to wireless sensor networks. It is based on methods from the field of sparse recovery, using Gaussian kernels as basis functions. After presenting the underlying physical process, a linear system model is developed out of it. For the estimation of its unknown parameters, solution approaches are presented. The approach is verified by means of numerical simulations. Furthermore, we indicate ways to perform a distributed estimation of the parameters within the network.