Base Band Data for Testing Interference Mitigation Algorithms

Jon F. Bell, Peter J. Hall, Warwick E. Wilson, Robert
J. Sault,
\\ Rick J. Smegal, Malcolm R. Smith, Willem van Straten, \\
Michael J. Kesteven, Richard H. Ferris, Frank H. Briggs, \\
Graham J. Carrad, Malcom W. Sinclair, Russell G. Gough, \\
John M. Sarkissian, John D. Bunton \& Matthew Bailes, PASA, 18 (1), in press.

---

Next Section: Conclusions
Title/Abstract Page: Base Band Data for
Previous Section: Reference Antenna
Contents Page: Volume 18, Number 1

---

Subsections

• Array data - ATCA at Narrabri
• Parkes 64m

---

Description of available data

Array data - ATCA at Narrabri

The first data set from CSIRO's Australia Telescope Compact Array (ATCA) at Narrabri (designated aieo in Table 2) included an OH source and interfering signals from a LEO satellite, MDS TV and GLONASS (Smegal & Wilson 1999). The data contains single polarisation 8MHz band widths, sampled using 4-bits for each of 6 antennas. A second Narrabri experiment (designated srtca in Table 2) collected more such data, but this time with dual polarisation, more carefully controlled sampler levels and a separate reference antenna. In this case 4MHz of band width was recorded with 4-bit dynamic range. Recordings included interfering signals from GLONASS, GPS, and microwave links, together with astronomical signals from an OH maser, a pulsar, and a continuum source. The reference receiver was bolted onto the backing structure of the eastern most antenna (antenna 1) (i.e., the unshadowed antenna closest to the transmitters on nearby Mt Dowe (east of the ATCA). The main reason for doing this was to (a) simplify the cabling required to connect the signals into the standard conversion system and (b) to make use of the antenna drive systems in pointing the reference receiver. While the reference receiver was in use, no astronomical signals were recorded from antenna 1. The outputs of the reference receiver were fed via heliax cable and plugged directly into the input of the 1400 MHz band RF module of the ATCA receiver conversion system (Gough & Graves 1999). The locations of the antennas and other systems parameters for the datasets are summarised in Table 3. In both cases, the data was recorded using three S2 recorders (Cannon et al. 1997), allowing a total recording rate of 384 Mb/s. The data was extracted from the S2 tapes using the s2tci system (Wietfeldt et al. 1998), demultiplexed and stored on DLT and CD media. Table 2 summarises the data collected. Sample spectra for a range of the data sets are shown in Figure 4a - 4f. These spectra were obtained by integrating several seconds of data in most cases. The mapping of the 4-bit levels is as follows:
In the aieo data set, there is one polarisation and one antenna per file which contains a simple stream of 4-bit numbers. For the srtca dataset there are 2 polarisations and 2 antennas per file. Each file contains four interleaved 4-bit data streams with each 16-bit word containing:

Parkes 64m

At Parkes the DLT based Caltech Parkes Swinburne recorder (CPSR) system was used (van Straten et al. 1999). A range of interferers (including point-to-point microwave links and the sun) and astronomical sources (pulsars) were recorded; details are summarised in Table 3 & 4. The first data set obtained contains dual polarisation for 2 beams from the multibeam system, the idea being to use one as the reference antenna for the other. However, much higher interference-to-noise ratio was obtained by using the reference receiver. The reference receiver was placed on the landing outside the top floor of the Parkes control room (schematically indicated in Figure 5a). From there, heliax cables carried the signals directly into the control room and down conversion equipment. Consequently the cable run is substantially shorter for the reference receiver than for the multibeam, leading to a significant delay mismatch between the two systems.

For both of the data sets collected, the resulting four signals were converted to base band. In the case of the multibeam however, base band 0-320 MHz signals were obtained from the regular multibeam equalisers. These signals were then passed through 5MHz low pass filters and fed directly into the CPSR, i.e., the normal CPSR quadrature sampling down converter was bypassed. The CPSR recording system was then used to 2-bit sample the data at 20 MHz (50 ns resolution) and write it to DLT tape in 1GB files. The data are therefore oversampled by a factor of 2. Four 2-bit samples (i.e., one for each of the 4 channels) are packed into each byte in the files on tape and on disk. Unfortunately it was only possible to record 2-bit data with the CPSR system, as the 4-bit recording mode was not working at the time. Example spectra in Figure 5 show the nature of some of the interfering signals.

---

Next Section: Conclusions
Title/Abstract Page: Base Band Data for
Previous Section: Reference Antenna
Contents Page: Volume 18, Number 1

---

© Copyright Astronomical Society of Australia 1997