Efficient Approximate Wordlength Optimization

  • Authors:
  • Paul D. Fiore

  • Affiliations:
  • MIT Lincoln Laboratory, Lexington

  • Venue:
  • IEEE Transactions on Computers
  • Year:
  • 2008

Quantified Score

Hi-index 14.99

Visualization

Abstract

In this paper, the problem of bounding the performance of good wordlength combinations for fixed-point digital signal processing flowgraphs is addressed. By formulating and solving an approximate optimization problem, a lower bounding curve on attainable cost/quality combinations is rapidly calculated. This curve and the associated wordlength combinations are useful in several situations, and can serve as starting points for real design searches. A detailed design example that utilizes these concepts is given.