Contents I. Introduction
In modern electronic environments, it is becoming increasingly important to deploy multiple agents with the autonomy capability of coordinating among themselves in conformance to pre-specified inter-agent constraints [1]. In one way or another, a large class of multiagent coordination problems can be described in a general constraint design framework as follows: Given a network of interacting agents with individual goals, and a set of inter-agent constraints that govern the high-level behavior of these interacting agents, synthesize local plans and coordination strategies, called coordination modules (CM's) henceforth, for every agent so as to guarantee that the agents can achieve their local goals and their coordinated behavior will conform to every inter-agent constraint without conflict or blocking. Conflict can occur either when the actions allowed by an agent are not allowed by another agent, or when the actions an agent needs to take in handling a constraint task are not concurrently allowed in handling another task. In either case, agents are prevented from reaching their design goals.
