Dissertation: Algorithms for Radio Interference Detection and Removal

Cover of my dissertationOn 22 June 2012 I defended my thesis on "Algorithms for Radio Interference Detection and Removal". The complete manuscript pdf can be downloaded here: Thesis A. R. Offringa.


  1. LOFAR fulfills its high expectations!
  2. Radio-astronomical observations in populated areas can achieve very high sensitivities.
  3. Continuous broadband transmitters, such as DAB stations, are currently the most destructive transmitters for performing low-frequency radio astronomy.
  4. Compared to conventional thresholding, the SumThreshold method is more accurate for detection of interference, and should be used for data from all sensitive radio telescopes. (Chapter 2)
  5. Morphological detection, especially the scale-invariant rank operator, is a good and fast additional step for finding interference. (Chapter 2)
  6. It is very likely that radio-frequency interference will not prevent a LOFAR detection of signals from the Epoch of Reionisation. (Chapter 6)
  7. While the introduced low-pass filters are still in its infancy, they appear to be an efficient technique for the attenuation of undesirable relics in the final sky maps. (Chapter 4)
  8. Sources of interference for LOFAR are approximately uniformly distributed on the surface of the Earth. The propagation of their signals can be well described by Hata's model. (Chapter 6)
  9. For determining the best location of a low-frequency radio telescope, the average condition of the ionosphere is more important than the amount of RFI.
  10. Astronomical research will benefit from collaboration with computer science.
  11. Research can be compared to ballroom dancing: it is important to be able to stand and move independently, but even better results can be achieved by combining forces.
  12. "By definition" is by definition not equal to "almost always", even though it is almost always used like that.

André Offringa