Microphone Array Research at CAIP 
  • Introduction
  • HMA
  • Sound Capture
  • Source Location
  • Sensor Placement
  • Array-Based Speech Recognition
  • Publications
  • References
  • People


    • Maintained by Chris Alvino.  Last updated 11/97. 

    Sensor Placement 

    There are two aspects of microphone array system performance: The ability of the system to locate and track sound sources, and the ability of the system to selectively capture desired sound sources. Both aspects of system performance are strongly affected by the spatial placement of microphone sensors. Research has been conducted on developing a method to optimize sensor placement based on geometry of the environment and assumed sound source behavior. The objective of the optimization is to obtain the greatest average system SNR using a specified number of sensors. 

    Sound source location may be implemented in a two step process: The first step estimates time delay of arrival (TDOA) for select microphone pairs. A modified version of the Omologo-Svaizer Cross-Power Spectrum Phase Method may be used to compute TDOA estimates. The second step uses the computed TDOA estimates in a least mean squares gradient descent search algorithm to compute a location estimate. The accuracy of the location estimate is dependent on aggregate sensor pair placement, orientation, and separation distance. The dependence of TDOA's reliability on these sensor parameters may be evaluated experimentally for a given acoustic environment. 

    The performance of sound source capture is also dependent on array geometry and the acoustical environment. Simple beamforming provides spatially selective gain in capture of a desired sound source. The shape and gain advantage of the capture region is highly influenced by the placement of array sensors. The performance of beamforming is degraded by reverberation due to the topology of the acoustic enclosure. Advanced techniques such as multiple beamforming and matched filtering are used to mitigate the effects of reverberation. 

    Research has been conducted towards evaluating array performance as described by the above mentioned metrics, for a given array configuration. An overall performance function could then be described based on these metrics. A framework for optimum placement of sensors under the practical considerations of possible sensor placement and potential location of sound sources may then be described.