Multi-Function Digital Signal Processing System for a 110-meter Radio Telescope

The QiTai radio Telescope (QTT) is under construction in northwest China. It will be equipped with multiple single-beam and Phased Array Feed (PAF) receivers to achieve continuous frequency coverage from 270 MHz to 30 GHz in the first stage. The receivers are located in separate places with different frequencies, bandwidths, and number of channels, which increases the design complexity of signal acquisition and transmission. The Ultra-WideBand (UWB) and PAF receivers will generate a large amount of data, and the signal processing algorithm for multi-object observation is complex and computational, which poses a great challenge to the data processing system. Based on the integrated, high-performance Radio Frequency System-on-Chip (RFSoC) technology, sampling the RF signals directly at the receiver front-end, a state-of-the-art and resource-shareable QTT signal acquisition and transmission scheme will be shown in this paper. To cope with the real-time processing of massive data, QTT will employ a multifunctional post-processing system based on High-Performance Computer clusters (HPC) that can handle a variety of scientific observation modes such as pulsars, transients, molecular lines, continuous and VLBI/baseband, etc. Many useful features such as Radio Frequency Interference (RFI) cancellation, commensal observation, and raw voltage data capture are also included in this system. The scalability of this heterogeneous architecture is excellent, and more channels or larger bandwidth receiver signals can be expended by increasing the number of RFSoC and HPC processing nodes.