ASPH 611 Term Project
A University of Calgary Department of Physics and Astronomy
Graduate Course in Radio Astronomy
Christy Bredeson, December 12, 2005
Simultaneous advances in radio technology and communications have caused a serious increase in the levels of interference seen at radio frequencies. Not only are radio technologies transmitting at specific frequencies, but they also tend to have power that leaks into adjacent frequencies for a broadband effect. Such advances have now made it impossible to obtain a set of astronomical data in the radio band completely free of interference. Thus, it is very important that adequate procedures be developed to identify and remove radio frequency interference (RFI) signals that pollute otherwise usable data.
Our radio telescope, which is trying to detect the HI line at about 1420 MHz, is not exempt from interference due to nearby sources. Although Canada has allocated a 27 MHz band for radio astronomy, as well as passive use by Earth exploration satellites and space research, RFI can still be seen at some key frequencies, occasionally even blocking one of the strongest signals in radio astronomy: the HI line itself! Large voltage spikes occur somewhat randomly, with some channels that are more affected than others, and can last either a short period of time or the duration of the observation. Although the removal of RFI is not necessarily critical for a spectral line observation, these spikes can greatly affect the scientific use of a data set, especially if they are on or near your frequency of interest.
One frequency in particular that proved to be troublesome during our observations was 1132 MHz. This is one of the many frequencies used by Calgary's airport radar and legally allocated to aeronautical radionavigation (the full range is from 960 MHz to 1164 MHz). Normally a frequency outside our range of frequencies of observation would not be a problem. However, by using a local oscillator frequency that was exactly 144 MHz above 1132 MHz, this frequency was a beat frequency of the system and was passed down through to the second mixer in the system, centered at 144 MHz or about 1420 MHz, very close to the HI line. Several solutions were considered to remedy this problem. A high pass filter was ordered to pass only frequencies higher than 1400 MHz, but it did not arrive in time to be useful for the project. A high pass filter was also designed for use in the system, but it did not quite work as expected. It added substantial noise to the system since it needed to be on the front end of the telescope's electronics and overall, did not work very well. In the end, we just did nothing and did our best to ignore this prominent RFI.
Each day, RFI was a problem in our observations. The primary RFI problem was being generated in the building underneath the telescope. ScienceB is shared with both Physics and Chemistry which both have high energy experiements running at the time of the observations. In particular, the NMR (Nuclear Magnetic Resonance) equipment emits sweaps of radio energy that heavliy interefered with our measurements.
Although we were still able to detect the HI line, the difficulity in disassociating the sky signals from the RFI was challenging. Significant improvement was made on the original data by post processing in software. Below is a comparision of the raw data and the RFI cleaned up data with some custom code written by Jeff Dever:
Last modified: 10:22 am July 17, 2014