Dynamic matching approach for interest management in distributed agent-based simulation

One key issue for distributed agent-based simulation is to maintain the shared environment state. The approach that distributes the environmental state to the relevant agents can overcome the performance bottleneck of the central way, while it needs to minimize irrelevant data transmission in network by using interest management. The performance of interest management strongly relies on the procedure that detecting relevant information between agents. This paper proposes an efficient matching approach which is good to deal with dynamic scenarios by limiting the matching computing only in the moving path. Our experiment results show that this matching approaching is suitable for the requirements of large-scale distributed agent-based simulation.