Abstract Details
| Name: Kaushal Buch Affiliation: GMRT, NCRA-TIFR Conference ID: ASI2015_657 Title : Real-time Radio Frequency Interference Mitigation for the GMRT Wide-band Digital Backend Authors and Co-Authors : AjithKumar B., GMRT, NCRA-TIFR Yashwant Gupta, NCRA-TIFR Abstract Type : Poster Abstract Category : Instrumentation and Techniques Abstract : The Giant Metrewave Radio Telescope (GMRT) is one of the most sensitive instruments for observing astrophysical phenomena at metre wavelengths. Currently, the GMRT is undergoing a major technological upgrade which is aimed at achieving almost seamless frequency coverage from 130 to 1450 MHz, with a maximum instantaneous bandwidth of 400 MHz, which will lead to a significant increase in the sensitivity and capabilities of the radio telescope. While the increase in receiver bandwidth would result in better signal-to-noise ratio, it also likely to encounter higher levels of man-made radio frequency interference (RFI). In order to achieve the expected sensitivity and dynamic range, it is necessary to mitigate the effects of RFI from the astronomical signal. This paper describes real-time filtering techniques for excision of broadband RFI from the Nyquist-sampled (800 MHz) digital time series of the received signal. This non-linear filter uses robust statistical estimation through Median Absolute Deviation which is followed by filtering of samples that are detected outside the robust threshold. This RFI filter is optimized for operation in real-time on the Field Programmable Gate Array (FPGA) board used for receiving the digitised signal. It is currently being tested with the new GMRT wideband backend (GWB). Initial test results show successful removal of broadband RFI and subsequent improvements in the signal-to-noise ratio. This paper describes different modes of operation of the filter and the test results in detail, and also outlines the plan for future development and use in regular observations. An introduction to the filtering of narrowband interference using the same technique operating on the spectral domain data is also provided, along with the test results from simulations. The broad plan for future development of real-time RFI filtering techniques for the upgraded GMRT is also outlined. |