Reports

A Distributed Signal Processing Facility for Speech Research

Polish, Nathaniel

An interactive digital voice laboratory facility has been developed to allow the sharing of expensive signal processing resources among large numbers of interactive display devices. The environment considered in this work is a facility in which there are a moderate number of heterogeneous systems with extensive floating point numerical capability. In this environment there are a number of voice I/O terminals with the capability to display and play voice in a variety of ways. Using shared me access and remote procedure calls, any of the voice workstations may have signal processing jobs performed on any of the available floating point processors without explicit knowledge of where the computation is being performed. The appearance to the user of the voice workstation is that of a workstation with enormous resources.

Subjects

Files

More About This Work

Academic Units
Computer Science
Publisher
Department of Computer Science, Columbia University
Series
Columbia University Computer Science Technical Reports, CUCS-386-88
Published Here
December 21, 2011