Interference response of adaptive optimal array filters

Abstract

When signal parameter estimates are inexact, one approach is to use multiple OAFs to extract individual signals from which signal estimates may be refined to give new filters. In temporal regions where only coherent interference is present, the (unwanted) filter output may be combined with gradient information to adapt signal parameter estimates and hence improve filtering. This work directly obtains the filter rejection response in terms of signal parameters so that an analytical expression for gradient may be derived. It is believed that an exact gradient expression will lead to improved adaptive filtering.<>