Abstract :
In recent years radioastronomical receiving systems have seen huge improvements in terms of bandwidth and back-end processing capabilities. As an example, the new K-band receiver that is currently under test at the Effelsberg 100-m radio telescope provides 8-GHz of instantaneous bandwidth and, using stacked FPGA-based FFT spectrometers, a total of more than 1 million spectral channels per polarization. While such new devices offer great potential to the modern observatory facilities, they also demand completely new data reduction software. On the one hand, the sheer amount of data produced needs sophisticated programming techniques to make use of parallel/distributed computing. One the other hand, for an accurate calibration, it must be taken into account that quantities like system temperature, Tsys, the intensity, Tcal, of calibration normals (noise diodes), and receiver and antenna gains are not constant over the large recorded bandwidth.
