A method of sidelobe cancellation using wavelet packets

Abstract

Classical methods of performing sidelobe cancellation include an approach whereby the main and auxiliary channels are divided into a prescribed number of contiguous subbands. This is known as a band-partitioned (BP) sidelobe canceler (SLC). A method of BP sidelobe cancellation using wavelet packets (WP) has been developed. The use of a wavelet packet decomposition is a simple, well-defined way to implement a paraunitary tree-structured filter bank. The performance of this method is compared to the method used in Lin et al. (2003), which is a special case of a uniform DFT filter bank. It is shown that the WP analysis filters provide better separation between adjacent narrowband channels, leading to an increase in performance of the BP SLC algorithm.