Daily Image
05-07-2007
LOFAR Core Station CS1 spectral occupancy statistics
| Submitter: | Albert Jan Boonstra and Harmen Jeurink |
| Description: | The figures above show LOFAR autocorrelation spectra of a single antenna, based on a 24 hour High Band Antenna observation. The spectrum band is allocated to land-mobile 0.1-10 W E.R.P. applications (cf. http://www.agentschap-telecom.nl/nfr/). The station signals were correlated by the LOFAR central computing facility (CEP, BlueGene/L), and integrated to one minute data points. Concerning time-continuous weak interferemce (RFI), these graphs show sensitivities which are about three to five orders of magnitude better than those of standard spectrum monitoring systems. The figures show that most of the RFI in this frequency band, at least on the ninth of April, is very weak and has a time-frequency occupancy of a few percent. This makes this band very suitable for LOFAR observations. The percentile statistics, shown in the upper figure, were derived from 1429 sorted data samples. These curves were corrected for the station subband filter shape, and for the slowly varying time dependence of the observed sky noise power. A residual fluctuation and a slope (partly caused by the spectral index of the sky background) are visible. After removing these residuals, both the corrected median and the 24 hour power average were derived. These are visible as black curves in the upper figure. An RFI detection threshold, depicted by the horizontal red line in the upper figure, was defined as two times the standard deviation of the observed 24 hour power average. Signals exceeding the threshold were considered RFI, and were re-plotted after subtracting the system noise. The system noise was defined as the average of the corrected median. The lower figure, which is a zoomed-out version of the upper, shows in blue the RFI power levels corrected for system noise. The horizontal parts of this curve indicate the corrected RFI power level detection limit. The blue curve lies a few dB above the theoretical detection limit, indicated by the lower red curve. This is because the spectral baseline ripple was not fully removed, leading to an increase of the estimated power standard deviation, and as a consequence to a higher threshold. |
| Copyright: | LOFAR |
| Tweet |  |