ATM network management based on distributed artificial intelligence architecture

In this paper, we describe a multi-agent architecture for Virtual Path (VP) management (i.e. bandwidth and restoration) in Asynchronous Transfer Mode (ATM) networks. The intention is not to propose new strategies, but to improve the management integrating these different management mechanisms by using distributed artificial intelligence. Major objectives are scalability and robustness of the system. Our approach uses two different multi-agent systems. The first one consists of simple and small reactive agents that control and monitor the network elements (nodes and virtual path). The second one is another multi-agent system of more complex agents that monitors and controls the previous ones, while it is also allowed to change the allocated resources.