Synthesis from Temporal Specifications Using Preferred Answer Set Programming

Abstract

We use extended answer set programming (ASP), a logic programming paradigm which allows for the defeat of conflicting rules, to check satisfiability of computation tree logic (CTL) temporal formulas via an intuitive translation. This translation, to the best of our knowledge the first of its kind for CTL, allows CTL reasoning with existing answer set solvers. Furthermore, we demonstrate how preferred ASP, where rules are ordered according to preference for satisfaction, can be used for synthesizing synchronization skeletons of processes in a concurrent program from a temporal specification. We argue that preferred ASP is put to good use since a preference order can be used to make explicit some of the decisions tableau algorithms make, e.g. declaratively specifying a preference for maximal concurrency makes synthesis more transparent and thus less error-prone.