Adaptive Beamforming for Audio Signal Acquisition
This chapter provides an overview of adaptive beamforming techniques for speech and audio signal acquisition. We review basic concepts of optimum adaptive antenna arrays and show how these methods may be applied to meet the requirements of audio signal pr
- PDF / 697,632 Bytes
- 40 Pages / 439.37 x 666.142 pts Page_size
- 71 Downloads / 214 Views
6.1
Introduction
Array processing techniques strive for extraction of maximum information from a propagating wave field using groups of sensors, which are located at distinct spatial locations. The array sensors transduce propagating waves into signals describing both a finite spatial and a temporal aperture. In accordance with temporal sampling which leads to the discrete time domain, spatial sampling by sensor arrays forms the discrete space domain. Thus, with sensor arrays, signal processing operates in a multidimensional space-time domain. The processor which combines temporal and spatial filtering using sensor arrays is called a beamformer. Many properties and techniques which are known from temporal finite impulse response (FIR) filtering directly translate to beamforming based on finite spatial apertures1 . Usually, FIR filters are placed in each of the sensor channels in order to obtain a beamformer with desired properties. Design methods for these filters can be classified according to two categories: (a) The FIR filters are designed independently of the statistics of the sensor data (data-independent beamformer). (b) The FIR filters are designed depending on known or estimated statistics of the sensor data to optimize the array response for the given wave-field characteristics (data-dependent beamformer) [2]. 1
See [2] for a tutorial about beamforming relating properties of temporal FIR filtering and space-time processing.
J. Benesty et al. (eds.), Adaptive Signal Processing © Springer-Verlag Berlin Heidelberg 2003
156
W. Herbordt, W. Kellermann
Generally, array signal processing is applied to detection and estimation problems when a desired signal is captured in the presence of interference and noise. Arrays play an important role in areas like (a) detection of the presence of signal sources, (b) estimation of temporal waveforms or spectral contents of signals, (c) estimation of directions-of-arrival (DOAs) or positions of multiple sources, and (d) focusing on specific spatial locations for transmission. Traditionally, they are used in such diverse fields as radar, sonar, transmission systems, seismology, or medical diagnosis and treatment. A new and emerging research field is given by arrays of microphones for space-time acoustic signal processing. Typical applications include hands-free acoustic human-machine interfaces for enabling acoustic telepresence as desired for audio-/video-conferencing, dialogue systems, computer games, command-and-control interfaces, dictation systems, highquality audio recordings, and other multimedia services. All these applications of microphone arrays have in common that they focus on speech and audio signal acquisition (estimation) in the presence of noise and interference. Obviously, spatio-temporal filtering is preferable over temporal filtering alone, because desired signal and interference often overlap in time and/or frequency but originate from different spatial coordinates. Spatial filtering by beamforming allows separation without distortion of the desired signal. In practic
Data Loading...
