Algorithms for Categoric Analysis of Interference in Low Frequency Radio Astronomy

Abstract

A study of the various types of terrestrial electromagnetic interferences categorized in terms of frequency, flux density, radio-direction, and statistics on a regular basis is essential at the radio telescope sites for understanding the radio environment viz. EMI/EMC, out of band/spurious emissions, for radio-astronomical frequency allocation and monitoring the growth of transmitters etc. within the region. This requires continuous monitoring of raw terrestrial spectrum data using spectrum monitoring systems and analyze the data collected over a period of time. At any instant of time, the raw spectrum data appears like many narrow band tall vertical lines standing over a relatively smooth wide band base-spectrum across the entire band. The statistical analysis for spectral occupancy gains in accuracy if somehow the narrow and broad band spectrums are separated out from the raw data and expressed in time-frequency-power space. Here we describe in details an algorithm for such an analysis. The algorithm first separates the broad band spectrum features from raw data in two steps, viz. chopping the narrow band peaks followed by a virtual low pass filtering process. The narrow band spectrum features are separated by comparing the broad band spectrum features with the raw spectrum data. Statistics of radio interference are then generated separately for each category. Analyzed results of some real interference spectrum data collected at a radio astronomy observatory site are also produced.