Post-correlation radio frequency interference classification methods

Abstract

We describe and compare several post-correlation radio frequency interference\nclassification methods. As data sizes of observations grow with new and\nimproved telescopes, the need for completely automated, robust methods for\nradio frequency interference mitigation is pressing. We investigated several\nclassification methods and find that, for the data sets we used, the most\naccurate among them is the SumThreshold method. This is a new method formed\nfrom a combination of existing techniques, including a new way of thresholding.\nThis iterative method estimates the astronomical signal by carrying out a\nsurface fit in the time-frequency plane. With a theoretical accuracy of 95%\nrecognition and an approximately 0.1% false probability rate in simple\nsimulated cases, the method is in practice as good as the human eye in finding\nRFI. In addition it is fast, robust, does not need a data model before it can\nbe executed and works in almost all configurations with its default parameters.\nThe method has been compared using simulated data with several other mitigation\ntechniques, including one based upon the singular value decomposition of the\ntime-frequency matrix, and has shown better results than the rest.\n