Challenger: a multi-agent system for distributed resource allocation

In this paper we introduce Challenger, a multiagent system that performs completely distributed resource allocation. Challenger consists of agents which individually manage local resources; these agents communicate with one another to share their resources (in this particular instance, CPU time) in an attempt to more e ciently utilize them. By endowing the agents with relatively simple behaviors which rely on only locally available information, desirable global system objectives can be obtained, such as minimization of mean job ow time. Challenger is similar to other market-based control systems in that the agents act as buyers and sellers in a marketplace, always trying to maximize their own utility. The results of several simulations of Challenger performing CPU load balancing in a network of computers are presented. The main contribution of this research is the addition of learning to the agents, which allows Challenger to perform better under a wider range of conditions than other systems for distributed processor allocation, such as Malone's Enterprise [Mal88].