Iterative spectrum analysis for pre-cleaning of narrow-band interference from radar data

Examples of pre-cleaning of ionograms are presented by using a new method of spectrum analysis of 512 complex data samples, independently in the period after each transmitter pulse. The exact amplitude and exact frequency of the four largest interference signals are determined in the partial spectrum and subtracted from the time-domain data. A new spectrum is created from the modified data of a single pulse period until the remaining spectral amplitude peaks are less than 10 dB above the running average spectral amplitude. After this procedure the total data are processed the standard way. In contrast to all known methods of filtering the time-domain data (to reduce the leakage caused by the finite data sampling length) and the "punching out" of spectral lines with large amplitude, this method has the least effect on the amplitude and phase of the wanted data.