UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

High cadence kurtosis based RFI excision for CHIME Mirhosseini, Arash


This thesis describes the real-time Radio Frequency Interference detection system for the Canadian Hydrogen Intensity Mapping Experiment (CHIME). CHIME is a transit radio telescope located at Dominion Radio Astrophysical Observatory (DRAO) in Penticton, BC, and it is originally designed to map the large-scale structures in the redshift range 0.8<z<2.5 by observing the 21-cm emission line of the neutral hydrogen atom. One of the common problems for astrophysical radio observations is Radio Frequency interference (RFI) from terrestrial sources such as TV stations, airplanes, cellphones, etc. RFI detection and mitigation is an essential part of any research in radio astronomy, because RFI contaminates the astrophysical data and reduces the sensitivity of the telescope to the faint sources. Since most of the RFI is non-Gaussian and lasts less than one second, we developed a real-time high cadence RFI excision system for CHIME which uses the fourth statistical moment (kurtosis) to detect non-gaussianity in the signal. \par In this thesis, I introduce the algorithms for kurtosis based RFI excision that we have used in CHIME. The algorithms were tested and the results were compared with each other. I also discuss the effect of truncation of the samples in CHIME correlator on the spectral kurtosis estimates. I show that truncation bias causes the RFI system to flag bright celestial sources. I derive a correction for the truncation bias with a polynomial fitting and a cubic spline interpolation. Moreover, I evaluate the quality of the CHIME data taken between May 2019 and September 2020. I find that the RFI excision system can mitigate many types of RFI by excising less than 20% of the data (on average), from intermittent RFI caused by satellites or airplanes to static RFI, especially between 400 MHz and 500 MHz.

Item Citations and Data


Attribution-NonCommercial-NoDerivatives 4.0 International