Multithreaded constraint programming: A hybrid approach

Although powerful and robust, constraint programming often reduces to a backtracking constrained search. Consequently, the search time for many complex problems can be great. One solution is multithreaded par-allelization. Multithreaded parallelization can greatly improve search performance , not only increasing the size of solvable problems but also improving heuristic solutions that were previously limited by time. Presented is a hybrid approach to application-based multithreaded constraint programming. The approach uses threads to perform a parallel search that combines a best-rst approach and the standard backtracking approach. We illustrate the eeective-ness of the approach on the Set Covering Problem and report computational results of several large problem instances.