Abstract:

The authors proposed a multiagent cooperation method to solve heterogeneous resource allocation problem in a distributed way. When the problem is put into the architecture of distributed generalized assignment problem, the large problem is separated into many small ones and each agent is responsible for a part of the solution. By means of the knowledge from their previous communication searches, agents are able to predict how to deploy their future similar searches more efficiently. Therefore, if an agent can solve some parts well, similar searches will be sent to it. Using this method, we may avoid the shortcomings mentioned above in the traditional ways. The experiment shows that the proposed method is feasible.

Key words: heterogeneous resource allocation problem, multiagent system, distributed generalized assignment problem, heuristic algorithm