Browsing by Author "Wieringa, Siert"
Now showing 1 - 1 of 1
- Results Per Page
- Sort Options
- Incremental Satisfiability Solving and its Applications
School of Science | Doctoral dissertation (article-based)(2014) Wieringa, SiertThe propositional logic satisfiability problem (SAT) is a computationally hard decision problem. Despite its theoretical hardness, decision procedures for solving instances of this problem have become surprisingly efficient in recent years. These procedures, known as SAT solvers, are able to solve large instances originating from real-life problem domains, such as artificial intelligence and formal verification. Such real-life applications often require solving several related instances of SAT. Therefore, modern solvers posses an incremental interface that allows the input of sequences of incrementally encoded instances of SAT. When solving these instances sequentially the solver can reuse some of the information it has gathered across related consecutive instances. This dissertation contains six publications. The two focus areas of the combined work are incremental usage of SAT solvers, and the usage of parallelism in applications of SAT solvers. It is shown in this work that these two seemingly contradictory concepts form a natural combination. Moreover, this dissertations unifies, analyzes, and extends the results of the six publications, for example, by studying information propagation in incremental solvers through graphical visualizations. The concrete contributions made by the work in this dissertation include, but are not limited to: Improvements to algorithms for MUS finding, the use of graphical visualizations to understand information propagation in incremental solvers, asynchronous incremental solving, and concurrent clause strengthening.