From Wikipedia, the free encyclopedia

The DPLL/Davis-Putnam-Logemann-Loveland algorithm is a complete, backtracking-based algorithm for deciding the satisfiability of propositional logic formulae in conjunctive normal form, i.e. for solving the CNF-SAT problem.

It was introduced in 1962 by Martin Davis, Hilary Putnam, George Logemann and Donald W. Loveland, and is a refinement of the earlier Davis-Putnam algorithm, which is a resolution-based procedure developed by Davis and Putnam in 1960. Especially in older publications, the Davis-Logemann-Loveland algorithm is often referred to as the “Davis-Putnam method” or the “DP algorithm”. Other common names that maintain the distinction are DLL and DPLL.

DPLL is a highly efficient procedure, and after more than 40 years still forms the basis for most efficient complete SAT solvers, as well as for many theorem provers for fragments of first-order logic.

Contents 1 The algorithm 2 Current work 3 See also 4 References

The basic backtracking algorithm runs by choosing a literal, assigning a truth value to it, simplifying the formula and then recursively checking if the simplified formula is satisfiable; if this is the case, the original formula is satisfiable; otherwise, the same recursive check is done assuming the opposite truth value. This is known as the splitting rule, as it splits the problem into two simpler sub-problems. The simplification step essentially removes all clauses which become true under the assignment from the formula, and all literals that become false from the remaining clauses.

The DPLL algorithm enhances over the backtracking algorithm by the eager use of the following rules at each step:

Unit propagation 

If a clause is a unit clause, i.e. it contains only a single unassigned literal, this clause can only be satisfied by assigning the necessary value to make this literal true. Thus, no choice is necessary. In practice, this often leads to deterministic cascades of units, thus avoiding a large part of the naive search space.

Pure literal elimination 

If a propositional variable occurs with only one polarity in the formula, it is called pure. Pure literals can always be assigned in a way that makes all clauses containing them true. Thus, these clauses do not constrain the search anymore and can be deleted. While this optimization is part of the original DPLL algorithm, most current implementations omit it, as the effect for efficient implementations now is negligible or, due to the overhead for detecting purity, even negative.

Unsatisfiability of a given partial assignment is detected if one clause becomes empty, i.e. if all its variables have been assigned in a way that makes the corresponding literals false. Satisfiability of the formula is detected either when all variables are assigned without generating the empty clause, or, in modern implementations, if all clauses are satisfied. Unsatisfiability of the complete formula can only be detected after exhaustive search.

The DPLL algorithm can be summarized in the following pseudocode, where Φ is the CNF formula and μ is a partial truth assignment that is initially empty:

function DPLL(Φ, μ)
   if Φ=T 
       then return true;
   if Φ=F 
       then return false;
   if a unit clause (l) occurs in Φ
       then return DPLL(assign(l,Φ), μΛl);
   if a literal l occurs pure in Φ
       then return DPLL(assign(l,Φ), μΛl);
   l := choose-literal(Φ);
   return DPLL(assign(l,Φ), μΛl) OR DPLL(assign(NOT(l),Φ), μΛNOT(l));

In this pseudocode, assign(l,Φ) is a function that returns a formula that is obtained by replacing every occurrence of l with "true" and every occurrence of not l with "false" in the formula Φ, and simplifies the resulting formula. The pseudocode DPLL function only returns whether the final assignment satisfies the formula or not. In a real implementation, the satisfying assignment typically is also returned on success (this has been omitted for clarity).

The Davis-Logemann-Loveland algorithm depends on the choice of branching literal, which is the literal considered in the backtracking step. As a result, this is not exactly an algorithm, but rather a family of algorithms, one for each possible way of choosing the branching literal. Efficiency is strongly affected by the choice of the branching literal: there exist instances for which the running time is constant or exponential depending on the choice of the branching literals.

Current work on improving the algorithm has been done on three directions: defining different policies for choosing the branching literals; defining new data structures to make the algorithm faster, especially the part on unit propagation; and defining variants of the basic backtracking algorithm. The latter direction include non-chronological backtracking and clause learning. These refinements describe a method of backtracking after reaching a conflict clause which "learns" the root causes (assignments to variables) of the conflict in order to avoid reaching the same conflict again.

• Davis-Putnam algorithm
• Chaff algorithm
• Proof complexity

• Davis, Martin; Putnam, Hillary (1960). "A Computing Procedure for Quantification Theory". Journal of the ACM 7 (1): 201–215.

• Davis, Martin; Logemann, George, and Loveland, Donald (1962). "A Machine Program for Theorem Proving". Communications of the ACM 5 (7): 394–397.
• Ouyang, Ming (1998). "How Good Are Branching Rules in DPLL?". Discrete Applied Mathematics 89 (1–3): 281–286. DOI:10.1016/S0166-218X(98)00045-6.