Performance evaluation of transform selective interference suppression

Interference may degrade the performance of a communication or radar system, and use of an additional interference suppression (IS) algorithm may provide the means for reliable recovery of information from the received signal. This paper investigates the transform selective interference suppression algorithm (TSISA), which is an adaptive method for suppressing several types of interferences. The TSlSA is based on several parallel transforms, in which the most suitable transform domain is chosen. The selection is based on the compression gain metric, which defines how well the energy of the interference is concentrated in the particular transform domain. Considered interfering signal types are impulses, sinusoids and chirp signals. The simulation results confirm that the TSISA is capable of choosing the most suitable transform for the interference excision. In general, the stronger the interfering signal is, the more frequently the TSISA is able to choose the most suitable transform. The simulation results also show that the TSISA is able to excise all the considered interfering signals with a very reasonable complexity. In addition, the results indicate that the TSISA can reject impulse interference even if the proportion of corrupted samples is 80 % of all the samples.