Digital Backend Meeting: A Summary
By Bernd Klein, MPIfR Bonn, Member SOC
The first workshop on Digital Backend was held on 6 September at the Max-Planck-Institute for Radioastronomy in Bonn, Germany. This meeting was the first one in the series of upcoming engineering workshops sponsored and organized by the EU Consortium RadioNet within the Sixth Framework Program of the European Commission.
It is a pleasure for every conference summariser to review a day were every participant went well and many colleagues suggested to meet again in the near future. For me, the workshop gave a broad view of many topics in the field of digital backends including contributions of both practical as well as theoretical aspects.
The morning session was mainly dominated by spectrometers. It becomes clear that the classical auto-correlator is history and the next generation of spectrometers are based on Field Programmable Gate Array (FPGA) chips. Theses FPGAs, have got bigger, faster and better in the last years and allowing a continuously Fourier transform of time series data up to the Giga-Hertz sector today. Since FPGA chips are available as commercial of the shelf and very popular in the communication business, this new spectrometer technique is independent from customer designed (correlator-)chips and therefore allows a faster enhance in developing systems with larger bandwidth and increased spectral resolution. In contrast to the auto-correlation approach the direct Fourier transformation makes it possible to use multi-level analogue-to-digital converters (ADCs) resulting in an increased system sensitivity. In addition, the higher input dynamic range of the ADCs degrades the complexity of the IF-processing (no automatic gain control) and a separate measurement of the total power level is unnecessary. Talks by Jeff Mock, Dick Ferris and Edward Fry discussed the advantage in using polyphase filter banks (PFB) instead of a Fourier transforms in building spectrometers. Both the flatter pass band and the better stop band behaviour of the PFB predestine this approach to use in digital filter banks as well as in FX-correlators.
The afternoon session was concentrating on interferometers, like the upgrades at Plateau de Bure and the developments for the future projects ALMA and LOFAR. An excellent talk by Hans Hinteregger discussed the consequences of ADC non-linearities and gave also a good overview of the current status of available and planed high-speed ADCs.
From my point of view, the amazing realization of this workshop was that many contributions began with the words: "We use a fast ADC linked to a big FPGA and the processed data are piped to a PCI bus interface...". It becomes clear that many new backends are based on a very similar hardware approach -- a re-configurable design using FPGAs and that the digital engineers work changes more and more from the hardware development to the design of fast signal processing algorithms. Exactly this direction of novel backends shows that collaboration and communication between the participants are becoming increasingly important as digital technology speeds up. Maybe this workshop on Digital Backends could be a first step...
"Und dat", as they say in Cologne, "war et dann". I'd like to close by saying what a pleasure it was to participate in this workshop. Not only because of all the new developments in the Digital Backend section discussed, but also because of the good organisation of the whole day and the comfortable workshop dinner.
On behalf of all of the participants, I'd like to thank all the members of the Local Organising Committee for their great job.
And that's what the other participants said:
Tim Ikin, Jodrell Bank Observatory: "A gathering of global excellence! Very well planned and extremely useful. Thank you."
Stefan Müller, Universität Bern, Institut für angewandte Physik: "It was fine that there were people from user up to developer. For us it was good to meet people from RF Engines for our future work."
Jeff Mock, Berkeley, CA: "Excellent workshop, good talks and good conversation with others interested in the same thing. For the future, maybe more alcohol and I wish I knew about the telescope tour so that I could arrange travel plans."
André Gunst, ASTRON, NL: "Good initiative."
Guido Bartsch, FGAN-FHR/RWA, Wachtberg-Werthoven, Germany: "Sehr gute Organisation, insb. gute Themen/Vortragende ausgewählt; räumlich etwas beengt; dank Klimatisierung aber kein Problem..." (Very good organisation, good topics/speakers chosen. Room a little small but ok with the AC.)
Willem Baan, ASTRON, NL: "Constructive meeting. Collaboration and communication are becoming increasingly important as technology development speeds up."
Rob Millenaar, ASTRON, NL: "Make this a returning event. Rotating host/venue."
Guy Kenfuck, Observatoire de Nançay: "Good organization. We wish to do the same meeting the next year."
Jean-Michel Martin, Observatoire de Paris: "Among the ways to continue the discussions, we suggest that web pages links and participant (and their interested colleagues) email addresses could be posted on the Radionet Engineering Forum web site."
Christian Monstein, ETHZ Institute of Astronomy: "Very useful meeting! I suggest to repeat it in 1...2 years, since hardware and software changes tremendously. Exponential progress expected..."
Hansueli Meyer, ETHZ Institute of Astronomy: "Very good to have an international backend meeting in Europe. Would be great to have it next year again!"
Guntars Balodis, Ventspils International Radioastronomy Center: "Everything looks OK."
Paul Burgess, Jodrell Bank Observatory: "Very comprehensive with many items of interest. Well organised and presented."
Philippe Chavatte, IRAM: "A good idea and a success!"
Jouko Ritakari, Metsähovi Radio Observatory: "Good idea, many people seem to be working with the same projects and problems."
Axel Murk,Universität Bern, Institut für angewandte Physik: "Very well organised and interesting workshop. Many thanks!"
J.A. Lopez-Perez, OAN: "Congratulations for the quality of the presentations and for the organization."
Sergei Pogrebenko, JIVE: "Interesting and useful, good organized."
Bernd Klein, Max-Planck-Institut fuer Radioastronomie: "Everything looks good as long as you can do it in a digital manner!"