Opportunistic Synthesis in Reactive Games under Information Asymmetry

Abstract

Reactive synthesis [1]is used to synthesize a provablycorrect strategy (controller) with respect to a given Linear Temporal Logic (LTL) specification in a dynamic, potentially adversarial environment. The core idea of reactive synthesis is to model the interaction between a controlled agent (called the robot) and its dynamic and uncontrollable environment as a two-player turn-based zero-sum game. In particular, players’ Boolean payoffs are specified by temporal logic formulas [2]–a rich language that can expressive properties such as safety, liveness, stability, and temporally extended goals.