A Signal Dependent Compression Scheme for Radio Astronomy

Abstract

Radio astronomy requires the capture of cosmic radiation. The signals of interest can span large bandwidths and high dynamic ranges, and are often contaminated by powerful man-made radio frequency interference (RFI) from mobile communications, radar equipment and much more. Capturing such signals requires receiver systems that produce high bit level samples at tremendous data rates, much of which is wasted whenever the input signals are free of RFI and other transients. This thesis has explored methods of compressing the output streams of observatorydata recorders, by way of varying the bit level of samples depending on the momentary characteristics of the input signal, in order to reduce the effective output data rate and in turn more effectively utilize data storage equipment and transmission links. Simulations performed on exemplary, pre-recorded input signals consisting of noise contaminated by aircraft RFI have shown a potential data rate reduction of up to 70% compared to receiver systems used by some contemporary radio astronomy receiver systems.