The Abels System: Designing an Adaptable Interface for Linking Simulations

We have developed a framework for using software agent technology for linking distributed simulations and other data resources. The Agent-Based Environment for Linking Simulations (ABELS) framework uses software agents to coordinate distributed simulations, communicate data efficiently between simulations, and retrieve data from other sources such as sensors and datasets. ABELS allows simulations to enter and exit a global simulation “cloud” consisting of dynamically changing data and computational resources. Note that the networked resources may consist of other simulations, datasets, active probes, and sensors. The Agent-Based Environment for Linking Simulations (ABELS) provides a framework to facilitate the dynamic exchange of data between distributed simulations and other remote data resources. Specifically, the framework allows the formation of a dynamic “data and simulation cloud” that links a heterogeneous collection of networked resources. ABELS consists of three major components: user entities that serve as data producers and/or consumers, a brokering system for organizing and linking the various participants, and generic local agents that connect simulations and data resources to the cloud of participants. This paper describes the major redesign and implementation of the generic local agent, which serves as the adaptable interface between the user and the ABELS system. Each simulation in the cloud is designed independently with little or no knowledge of the other simulations in the cloud. While a simulation must be able to specify what resources it needs and what services it provides to the cloud, it does not need to know any specifics about the other simulations. The simulation also does not need to be written to a particular specification; instead, our agentbased framework provides the interface for linking the simulations and adapts to their needs.