Accession Number:

ADA288202

Title:

Numerical and Symbolic Algorithms for Application Specific Signal Processing.

Descriptive Note:

Semi-annual rept. 5 Apr-30 Oct 94,

Corporate Author:

MASSACHUSETTS INST OF TECH CAMBRIDGE RESEARCH LAB OF ELECTRONICS

Personal Author(s):

• Oppenheim, Alan V.

Report Date:

1994-11-21

Pagination or Media Count:

11.0

Abstract:

Work by Prof. Oppenheim and his collaborators is summarized here. The first step in one aspect of our research in the RASSP program involves evaluating the abilities and limitations of current DSP system design tools. We are concerned with what algorithms the tools support, and how the tool abilities might be expanded in the future to support emerging signal processing techniques. Major accomplishments in our recent approximate processing research include 1 conceptual and mathematical elaboration of approximate process- ing concepts, 2 formulation and implementation of a class of incremental refinement algorithms for the short-time Fourier transform, 3 illustration of design issues in the context of incremental refinement algorithms , 4 continued development of the IPUS-C software environment for supporting the design of application systems with embedded approximate signal processing capabilities, and 5 initiation of a project to build a software environment to support algorithm design for approximate signal processing.

Descriptors:

• *ALGORITHMS
• *SIGNAL PROCESSING
• *SOFTWARE ENGINEERING
• COMPUTER PROGRAMS
• FRACTALS
• AVIONICS
• REENTRY VEHICLES
• COMPUTERIZED SIMULATION
• INFRARED DETECTION
• FOURIER TRANSFORMATION
• PROGRAMMING LANGUAGES

Subject Categories:

• Computer Programming and Software
• Miscellaneous Detection and Detectors
• Numerical Mathematics

Distribution Statement:

APPROVED FOR PUBLIC RELEASE