Computer Science – Computational Complexity
Scientific paper
2004-06-02
Computer Science
Computational Complexity
Scientific paper
In 1978, Schaefer proved his famous dichotomy theorem for generalized satisfiability problems. He defined an infinite number of propositional satisfiability problems, showed that all these problems are either in P or NP-complete, and gave a simple criterion to determine which of the two cases holds. This result is surprising in light of Ladner's theorem, which implies that there are an infinite number of complexity classes between P and NP-complete (under the assumption that P is not equal to NP). Schaefer also stated a dichotomy theorem for quantified generalized Boolean formulas, but this theorem was only recently proven by Creignou, Khanna, and Sudan, and independently by Dalmau: Determining truth of quantified Boolean formulas is either PSPACE-complete or in P. This paper looks at alternation-bounded quantified generalized Boolean formulas. In their unrestricted forms, these problems are the canonical problems complete for the levels of the polynomial hierarchy. In this paper, we prove dichotomy theorems for alternation-bounded quantified generalized Boolean formulas, by showing that these problems are either $\Sigma_i^p$-complete or in P, and we give a simple criterion to determine which of the two cases holds. This is the first result that obtains dichotomy for an infinite number of classes at once.
No associations
LandOfFree
Dichotomy Theorems for Alternation-Bounded Quantified Boolean Formulas does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Dichotomy Theorems for Alternation-Bounded Quantified Boolean Formulas, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Dichotomy Theorems for Alternation-Bounded Quantified Boolean Formulas will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-618051