Adding Branching Temporal Dimension to Qualitative Coalitional Games with Preference

Abstract

Qualitative Coalitional Games (QCGs), as a variation of coalitional games, is to investigate agents’ strategies and behaviors in cooperating games. Each agent has a set of goals as its desires and will be satisfied if at least one of its desires is achieved by executing some strategies in a coalition, otherwise be unsatisfied. After introducing QCGs, we add preference to QCGs framework to enable that every agent has the ability to join the best coalition for achieving its preferences goals (QCGPs). In order to make a formal description and reason about repeated coalitional games, the paper will study Branching Temporal Qualitative Coalitional Games with preferences. Computational Tree Logic (CTL) is used for QCGPs with complete axiomatisation of it, denoted as CTQCGPs. Further more, this paper analyses the expression power, complexity, and some characteristic of CTQCGPs