Adaptive Filtering Approach to Multichannel Audio Reproduction Systems

One of the research areas of the Signal Processing group at the Eindhoven University of Technology is in adaptive array signal processing. It is concerned with the application of adaptive signal processing techniques to the outputs of a spatially distributed array of sensors. The power of these techniques lies in the possibility of achieving improvements in directivity and measured Signalto-Noise Ratio (SNR) as compared to that attainable with a single input sensor. The spatial dimensions introduced by processing a distributed array of sensors allow directional discrimination of signals against interference. Sensor arrays are used in many applications including radar, sonar, seismology and tomography [5]. Most of these applications use the fact that the used signals are band limited and that the sources are so far away that at the sensors they can be treated as a flat wavefront. In many (certainly ’in-the-house’) audio applications these narrowband and far field assumptions certainly do not hold. Mianly due to the vast increase of available computational power, commercial applications of array processing advanced audio systems have become feasible in recent years. Recently two PhD thesis [4, 11] have been successfully completed in the Signal Processing group in which adaptive array processing techniques were applied to audio applications. The work of [11] was focused around a part of the