FAST Digital Backend and Algorithm Simulation for Broadband Million Channel Spectrometer

Xia ZHANG,Xin-ying YU,Ran DUAN,Di LI,Jie HAO,Cheng-jin JIN
DOI: https://doi.org/10.3969/j.issn.1000-8349.2016.02.08
2016-01-01
Abstract:The National Astronomical Observatories of the Chinese Academy of Sciences (NAOC) will soon finish the Five-hundred-meter Aperture Spherical Radio Telescope (FAST), which will be the most sensitive single-dish radio telescope in the low frequency radio bands between 70 MHz and 3 GHz. To take advantage of its giant aperture, the ongoing self-developing brand-new broadband receiver China Reconfigurable ANalog-Digital backEnd (CRANE) is expected to ensure that it would achieve the best possible overall performance. <br> In hardware, CRANE includes developed parts — FAST Digital Backend (FDB), 3 Gbits/s, 12-bit Analog-to-Digital Converter (FAST3212ADC) — and developing part —Analog Front-end Board (AFB). In its firmware, FPGA-based digital signal processing al-gorithms such as Digital Down Converter (DDC) and Fast Fourier Transform(FFT) will be applied to realize channelization on FDB. CRANE plans to build million channel spectrome-ter which measures a frequency precisely on FDB. We utilize two cascaded FFT to optimize the FFT for large-scale operation. However, the resolution of the first FFT unit would cause the distortion of the output results. The simulation on Matlab shows that final frequency can be restored.
What problem does this paper attempt to address?