Process estimation and moving source detection in 2-D diffusion processes by scheduling of sensor networks

In this note, the problem of source detection for a diffusion process in a 2-D spatial domain is considered. An integrated state estimation and source detection scheme is proposed utilizing a sensor network. The sensing devices are assumed to measure spatial gradients as opposed to concentrations as this enhances the source-detecting abilities of the sensor network. Included in this scheme is a sensor scheduling policy that during a time interval employs a subset of the available sensors while the remaining ones are kept in a "sleep/idle mode". Numerical studies are included to support the proposed scheme.