High Time-Resolution Radio Frequency Interference and Single Pulse Pulsar and FRB Detection using Machine Learning Semantic Segmentation
Abstract
Radio frequency interference (RFI) is a problem that plagues nearly every radio astronomy observation. With wider bandwidth receivers and increasingly saturated RFI environments, there is a growing need for active, real-time RFI excision. We present a generalized machine learning approach to label each sample in coarse-channelized, frequency-space data as containing RFI. We also expand the method to detect time-variable astrophysical sources, such as fast radio bursts or bright pulsars. Using this method, the unwanted data can be differentiated for removal while simultaneously detecting rare astronomical signals. We discuss ways that this method could be optimized to run in real-time at microsecond time resolution using high performance, commercial-off-the-shelf computing hardware.
- Publication:
-
American Astronomical Society Meeting Abstracts #235
- Pub Date:
- January 2020
- Bibcode:
- 2020AAS...23510930H