Abstract:
Current telecommunication network management systems rely extensively on human intervention. They are also prone to fundamental changes as the managed network evolves. These two attributes, combined with the growing complexity of networks and services, make the cost of network management very high. In recent years, we have witnessed the emergence of artificial intelligence applications. Some are aimed at the creation of autonomic network management systems. This paper offers a novel approach to the design of a network management system that incorporates intelligent agents. As a benchmark to this model, we use two approaches most widely in use in network management systems today. The focus of this paper is on synchronization issues, service discovery and policy enforcement.
