Tagged BDDs: Combining reduction rules from different decision diagram types

Abstract

Binary decision diagrams are fundamental data structures in discrete mathematics, electrical engineering and computer science. Many different variations of binary decision diagrams exist, in particular variations that employ different reduction rules. For some applications, such as on-the-fly state space exploration, multiple reduction rules are beneficial to minimize the size of the involved graphs. We propose tagged binary decision diagrams, an edge-based approach that allows to use two reduction rules simultaneously. Experimental evaluations demonstrate that on-the-fly state space exploration is an order of magnitude faster using tagged binary decision diagrams compared to traditional binary decision diagrams.