Self-adjusting statistical noise analyzer with interference suppression

Abstract

Signals are processed along two paths, one determining the optimum freque band from which atmospheric noise will be analyzed and the other which performs statistical analysis of the frequency band found to be optimum. The band-setting path may include broadband filtering to limit a search to a desired band. A fast Fourier transform (FFT) sorts the analog-to-digitally converted signals into consecutive frequency bands identified herein as frequency bins. The magnitude of the signals falling within each frequency bin is determined and the average frequency bin magnitude is ascertained. Broadband peaks are searched by noting which frequency bin magnitudes are a predetermined amount over the average frequency bin magnitude. These frequency bins are 'marked' and the bins are then searched to determine the longest series of frequency bins between marks, thereby determining the widest frequency band bounded by broadband signals. In the statistical path, band path filtering is used to eliminate the broadband signals by setting the lower and upper cut-off frequencies according to the optimum frequency band found. The analog to digitally converted signals are fast Fourier transformed and the magnitude of the frequency bins are determined. Harmonics are excised through calculating the average magnitude of the frequency bins and normalizing any frequency bin with a magnitude greater than a predetermined amount over the average frequency bin magnitude.