Using Response Probability to Build System Redundancy in Multiagent Systems

This work shows that a simple response probability model may be used in decentralized multiagent systems to dynamically build and maintain redundancy in terms of agent experience. We model this system theoretically and use this model to provide guidance on how to choose response probability values that ensure that task needs are met and that ensure the a given level of redundancy is achieved.