Abstract
An adaptive method for suppressing interference with an unknown model is presented. The method is based on computation of several transforms in parallel and selecting the most suitable one depending on a priori unknown the interference type based on a compression gain (CG) metric, which is related to the interference concentration in the transform domain. It is also possible to estimate the a priori parameters required by some transforms. The method can be expanded to utilize an arbitrary number of transforms. It is shown through computer simulations that the CG is a good metric for choosing the proper domain for interference excision.