Current Research Focus: Sampling and Reconstruction of Diffusion Fields

  • Source and sensor configuration within the monitored region.
  • Summary

    Numerous biological and physical phenomena are modelled by the diffusion equation: temperature variation in fluids, disease epidemic dynamics, population dispersion and biochemical substance release are typical examples of such phenomena. Sensor networks (green 'o' and red 'o') are becoming increasingly prevalent for the monitoring of these physical phenomena. In such a setting we have access only to spatiotemporal samples of the field.

    My research interest focusses on estimating the sources (blue '+') of these fields, as well as reconstructing the entire field, given statiotemporal samples of the field obtained using sensor networks.

  • Real Temperature Data Experiments

  • Spatiotemporal samples of the diffusion field is obtained using nine sensors arbitrarily distributed over the region of interest. The sampling interval is 0.52 seconds. This video visualizes the evolution of the sensor measurements in space and time.

    Download video: MP4 format | Ogg format | WebM format

  • Measurements of two monitoring sensors obtained by two different sensors. The top plot shows the temporal samples of a sensor much closer to the source than the bottom plot sensor measurements. Indicating the instant of source activation is the dotted vertical line in each plot.
  • Related publications: [Murray-Bruce & Dragotti 2015] and [Murray-Bruce & Dragotti 2014a] and [Murray-Bruce & Dragotti 2014b].