A Strong Local Consistency for Constraint Satisfaction

Filtering techniques are essential to efficiently look for a solution in a constraint network (CN). However, for a long time it has been considered that to efficiently reduce the search space, the best choice is the limited local consistency achieved by forward checking [15, 17]. However, more recent works [18, 4, 16] show that maintaining arc consistency (which is a more pruningful local consistency) during search outperforms forward checking on hard and large constraint networks. In this paper, we show that maintaining a local consistency stronger than arc consistency during search can be advantageous. According to the comparison of the local consistencies more pruningful than arc consistency that can be used on large CNs in [9], Max-restricted path consistency (Max-RPC, [7]) is one of the most promising local consistencies. We propose a new local consistency, called Max-RPCEn, that is stronger than MaxRPC and that has almost the same cpu time require-