Blind Wideband Beamforming
Wei Liu, Stephan Weiss · 2010
This chapter focuses on a different class of adaptive beamformers, which is based on the concept of blind source separation (BSS). It gives a review of the BSS technique, with a kurtosis-based blind source extraction (BSE) algorithm as an example. BSS is a technique to recover the original sources from all kinds of their mixtures, without the knowledge of the mixing process and the sources themselves. For the linear mixing problem, it can be divided into two categories: instantaneous mixing and convolutive mixing, both of which can be found in the context of array signal processing. The chapter shows that for narrowband signals, one can have instantaneous mixing problem, while for wideband signals the problem is convolutive in nature. The blind wideband beamforming problem can be solved by existing convolutive BSS methods, although the computational complexity of such a system would be high even for a modest number of sensors. Controlled Vocabulary Terms array signal processing; blind source separation; mixing; narrowband; wideband