A Parallel-Oriented Language for Modeling Constraint-Based Solvers

This paper presents the Parallel-Oriented Solver Language (POSL, pronounced " puzzle "): a new framework to build interconnected meta-heuristic based solvers working in parallel. The goal of this work is to obtain a framework to easily build solvers and reduce the developing effort, by providing a mechanism for reusing code from other solvers. The novelty of this approach lies in looking at solver as a set of components with specific goals, written in a parallel-oriented language based on operators. An interesting advantage of POSL is the possibility to share not only information, but also behaviors, allowing solvers modifications. POSL allows solver's components to be transferred and executed by other solvers. It provides an additional layer to dynamically define the connectivity between the solvers. The implementation of POSL remains a work in progress, therefore this paper will focus on POSL concepts only.