Centralized Matchmaking for Minimal Agents

K. Bertels, N. Panchanathan, S. Vassiliadis, and B.P. Ebrahimi (The Netherlands)


: simulation, minimal agents, task allocation, market, match making.


In this paper we propose a simple though efficient mech anism to enable distributed processing nodes to commu nicate with each other and to dynamically rebalance the workload among them. We discuss a centralized matching mechanism and investigate different matching functions, varying in the amount of information that is taken into ac count. It is shown that the simplest mechanism, called FirstMatch, results in the fastest matchmaking achieving matching rates of 99%. We also show that this time effi ciency does not have a cost in terms of efficient task allo cation and resource utilization.

Important Links:

Go Back