Effective Variants of the Max-Sum Algorithm for Radar Coordination and Scheduling

Solving a coordination problem in a decentralized environment requires a large amount of resources and thus exploiting the innate system structure and external information as much as possible is necessary for such a problem to be solved in a computationally effective manner. This work proposes new techniques for saving communication and computational resources when solving distributed constraint optimization problems using the Max-Sum algorithm in an environment where system hardware resources are clustered. These techniques facilitate effective problem solving through the use of a pre-computed policy and two phase propagation on Max-Sum algorithm, one inside the clustered resources and one among clustered resources. This approach shows equivalent quality to the standard Max-Sum algorithm while reducing communication requirements on average by 50% and computation resources by 5 to 30% depending on the specific problem instance. These experiments were performed in a realistic setting involving the scheduling of a network of as many as 192 radars in 48 clusters.