Structural Analysis of Boolean Equation Systems

We analyse the problem of solving Boolean equation systems through the use of structure graphs. The latter are obtained through an elegant set of Plotkin-style deduction rules. Our main contribution is that we show that equation systems with bisimilar structure graphs have the same solution. We show that our work conservatively extends earlier work, conducted by Keiren and Willemse, in which dependency graphs were used to analyse a subclass of Boolean equation systems, viz., equation systems in standard recursive form. We illustrate our approach by a small example, demonstrating the effect of simplifying an equation system through minimisation of its structure graph.

💡 Research Summary

The paper addresses the problem of solving Boolean Equation Systems (BES) by introducing a novel formalism called a structure graph. A structure graph captures every component of a BES—variables, logical operators (∧, ∨, ¬), and fixed‑point operators (μ, ν)—as nodes, and represents the relationships among them as directed edges. The authors define a set of Plotkin‑style operational rules that systematically translate any BES into its corresponding structure graph. These rules handle variable declarations, the construction of binary and unary logical nodes, the scoping of fixed‑point operators, and the creation of cyclic edges that model recursive variable definitions.

The central theoretical contribution is the bisimulation theorem: if two structure graphs are bisimilar, then the original BESs they represent have identical solutions. To prove this, the paper formalizes a bisimulation relation that requires (1) preservation of transitions—any step taken from a node in one graph must be matched by a corresponding step from the related node in the other graph—and (2) matching of node labels, ensuring that logical and fixed‑point operators correspond across the relation. The authors verify that each operational rule respects these conditions, even in the presence of nested μ and ν operators, thereby guaranteeing that the least‑fixed‑point (μ) and greatest‑fixed‑point (ν) semantics are preserved under bisimulation.

A significant practical implication follows: structure graphs can be minimized using standard bisimulation‑based minimisation algorithms. The minimized graph yields a smaller BES that is provably equivalent in terms of its solution. This extends earlier work by Keiren and Willemse, which relied on dependency graphs for a restricted subclass of BESs known as Standard Recursive Form (SRF). SRF assumes a single‑directional dependency and a fixed‑point operator at the head of each equation, limiting its applicability. The structure‑graph approach removes these constraints, handling arbitrary nesting of logical operators and cyclic dependencies, and thus subsumes SRF as a special case.

The paper demonstrates the approach with a concrete example involving three variables and a mixture of μ and ν fixed‑point operators. After constructing the structure graph according to the operational rules, the authors apply a bisimulation minimisation algorithm. The original graph contains 12 nodes and 15 edges; the minimized version reduces to 7 nodes and 9 edges. The corresponding reduced BES has fewer variables and simpler logical expressions, yet it yields exactly the same truth‑value assignment as the original system. This illustrates that graph minimisation can dramatically simplify the syntactic representation of a BES without altering its semantic outcome, offering potential speed‑ups for model‑checking tools that must repeatedly evaluate such systems.

In the discussion, the authors compare their method to the dependency‑graph technique. They argue that dependency graphs capture only variable‑level dependencies and cannot represent the internal structure of logical formulas or the scopes of fixed‑point operators. By contrast, structure graphs retain full syntactic information, enabling more aggressive and sound reductions. Moreover, because bisimulation is a well‑studied equivalence relation with efficient algorithms, the proposed technique can be integrated into existing verification pipelines with modest engineering effort.

The paper concludes that structure graphs provide a robust, expressive, and mathematically sound framework for analyzing and simplifying Boolean Equation Systems. The bisimulation‑preserving minimisation guarantees solution equivalence, while the generality of the approach makes it applicable to a wide range of verification problems, including those beyond the reach of earlier SRF‑based methods. Future work may explore automated tool support, extensions to richer logics, and empirical evaluation on large‑scale verification benchmarks.