Abstract:Symmetries have been exploited successfully within the realms of SAT and QBF to improve solver performance in practical applications and to devise more powerful proof systems. As a first step towards extending these advancements to the class of dependency quantified Boolean formulas (DQBFs), which generalize QBF by allowing more nuanced variable dependencies, this work develops a comprehensive theory to characterize symmetries for DQBFs. We also introduce the notion of symmetry breakers of DQBFs, along with a concrete construction, and discuss how to detect DQBF symmetries algorithmically using a graph-based approach.
What problem does this paper attempt to address?
This paper attempts to solve the symmetry problem in Dependency Quantified Boolean Formulas (DQBFs) in order to improve solver performance and design more powerful proof systems. Specifically, the main objectives of the paper include:
1. **Develop a symmetry theory for DQBF**: The paper aims to establish a comprehensive symmetry - theoretical framework for DQBFs, which will help in understanding the structure and properties of these formulas. DQBFs are a class of logical formulas that allow existential quantifier variables to depend on a specific subset of universal quantifier variables, thereby being able to express more complex dependency relationships.
2. **Introduce symmetry breakers**: The paper introduces the concept of symmetry breakers for DQBFs and provides a specific construction method. Symmetry breakers are additional constraints that can break the symmetry of a formula, thus guiding the solver to avoid equivalent parts of the search space and accelerate the solving process.
3. **Propose an algorithm for detecting symmetry**: The paper discusses how to use graph - based methods to algorithmically detect symmetry in DQBFs. This method can help identify and utilize symmetry, thereby improving the solving efficiency.
### Specific Problem Description
DQBFs are a generalization of QBFs, allowing non - linear variable dependency relationships. Compared with QBFs, DQBFs have stronger expressive power, but also bring higher computational complexity (NEXP - TIME - complete). Therefore, studying the symmetry of DQBFs is crucial for improving their solving efficiency.
### Main Contributions
- **Symmetry Classification**: The paper distinguishes between two types of symmetry: syntactic symmetry and semantic symmetry. Syntactic symmetry involves transformations of the formula itself, while semantic symmetry involves transformations of the solutions of the formula.
- **Symmetry Breakers**: The paper proposes the concept of symmetry breakers and provides a specific construction method for breaking the symmetry in the formula, thereby guiding the solver to avoid equivalent parts of the search space.
- **Algorithm Detection**: The paper describes a graph - theory - based method for detecting symmetry in DQBFs, which provides an effective tool for practical applications.
Through these works, the paper provides an important theoretical basis and practical tools for the research and solution of DQBFs, and is expected to play an important role in fields such as verification, synthesis, and software - hardware engineering.
### Example
For example, consider the following DQBF:
\[
\forall x_1, x_2 \exists y_1(\{x_1\}), y_2(\{x_2\}). (\neg x_1 \rightarrow y_1) \land (x_2 \lor y_2)
\]
This formula cannot be represented by a standard QBF because its quantifier dependency relationship is non - linear. Through symmetry analysis, we can find that permutations of certain variables will not change the truth value of the formula, and thus these symmetries can be used to optimize the solving process.
In summary, this paper provides new ideas and methods for improving the solving efficiency by developing a symmetry theory for DQBFs.