A Probabilistic Approach to Resource Allocation in Distributed Fusion Systems

Abstract

In complex multi-agent fusion systems resource conflicts are very likely to occur. We propose an algorithm that determines the optimal sensing resource to fusion task assignment, based on the entropy change criterion. By exploiting the Bayesian network framework and the structure of our agent network the algorithm operates in a distributed manner by combining descriptions of local fusion models in an efficient way which provides significant advantages over centralized approaches.