A design lab for statistical signal processing

by Edward A. Lee

in Proceedings of IEEE Conference on Acoustics, Speech, and Signal Processing,
ICASSP-92, Vol. 4, pp. 81-84, San Francisco, CA, March, 1992.

ABSTRACT

In the spring of 1991 a software laboratory was added to the graduate statistical signal processing class at Berkeley. The emphasis of this lab was on high-level experimentation with signal processing algorithms. A separate course on design methodology for signal processing covers VLSI design and programmable digital signal processors (DSPs), so this particular lab steered clear of these issues. A graphical block diagram programming environment developed at Berkeley (called Ptolemy) was used on a network of DEC workstations. Commercial software such as Comdisco's SPW system could have been used as well. The students were assigned a sequence of six experiments and given two weeks to complete each one. The design of the experiments in view of the teaching objective and effectiveness are discussed.