Symmetry detection and dynamic variable ordering of decision diagrams

  • Authors:
  • Shipra Panda;Fabio Somenzi;Bernard F. Plessier

  • Affiliations:
  • Dept. of Electrical and Computer Engineering, University of Colorado at Boulder;Dept. of Electrical and Computer Engineering, University of Colorado at Boulder;Motorola Inc., Austin, TX

  • Venue:
  • ICCAD '94 Proceedings of the 1994 IEEE/ACM international conference on Computer-aided design
  • Year:
  • 1994

Quantified Score

Hi-index 0.01

Visualization

Abstract

Knowing that some variables are symmetric in a function has numerous applications; in particular, it can help produce better variable orders for Binary Decision Diagrams (BDDs) and related data structures (e.g., Algebraic Decision Diagrams). It has been conjectured that there always exists an optimum order for a BBD wherein symmetric variables are contiguous. We propose a new algorithm for the detection of symmetries, based on dynamic reordering, and we study its interaction with the reordering algorithm itself. We show that combining sifting with an efficient symmetry check for contiguous variables results in the fastest symmetry detection algorithm reported to date and produces better variable orders for many BDDs. The overhead on the sifting algorithm is negligible.