Matchmaking in the Abels System for Linking Distributed Simulations

Large-scale simulations often need dynamic access to heterogeneous data resources such as sensors, databases, or other simulations. The Agent-Based Environment for Linking Simulations (ABELS) is designed to facilitate the dynamic formation of a “cloud” of independent simulations and other data resources for the exchange of information. Participants in the data and simulation cloud join and exit the cloud as needed and have no prior knowledge of the other cloud participants. The formation of the cloud is achieved using a distributed brokering system that matches data consumers in the cloud with appropriate data producers, based on registration information submitted by the various participants in the cloud. This paper describes in detail the process used to match and rank prospective data producers for a given data consumer. The Agent-Based Environment for Linking Simulations (ABELS) system is designed to facilitate the rapid formation of a distributed “cloud” of autonomous data resources (Mills-Tettey et al. 2002; Mills-Tettey and Wilson 2003a; Mills-Tettey and Wilson 2003b; Murphy et al. 2003; Wilson et al. 2001; Wilson et al. 2002; Wilson et al. 2003). Individual cloud participants join and exit the data and simulation cloud as needed and have no prior knowledge of the other cloud participants. A distributed brokering system is used to match data producers to consumers and initiate communication between cloud participants, but it does not control the independently-designed participants in any way. Each participant in an ABELS cloud is responsible for determining what resources it makes available to the cloud and for describing those resources accurately.