Bound Consistencies for the discrete CSP

Many works in the area of Constraint Programming have focused on inference, and more precisely, on filtering methods based on properties of constraint networks. Such properties are called domain filtering consistencies when they allow removing some inconsistent values from the domains of variables, and bound consistencies when they focus on bounds of domains. In this paper, we study the relationship between consistencies introduced with respect to discrete and continuous constraint networks, and experiment the effectiveness of exploiting bound consistencies on discrete instances.