Advanced ordering and manipulation techniques for binary decision diagrams

Abstract

Heuristics leading to improved ordering computation for binary decision diagrams (BDDs) are given. An initial step, based on the topology of the network, generates a hierarchical variable ordering. This initial result is further refined by incremental manipulation governed by the stochastic evolution technique. A new property of BDDs is introduced as well, which accelerates commonly used operations. Experimental results are presented. Binary decision diagrams (BDDs) have been known for a long time as a means to represent the structure of switching functions, but only recently their use emerged in the implementation of design automation tools for digital circuits.<>