Addressing Interoperability in Network Management with Intelligent Agents

Both scholars and practitioners give increasing attention to the problems arising from the unpredictable nature and dynamism of modern heterogeneous networks. New means are explored that tackle consequent non-deterministic behaviors. In this paper, we argue that the Intelligent Agents constitute an attractive technology to address the issues challenging the managers of such networks. Management systems can be designed using agent-based engineering with a focus on the knowledge of the domain. Artificial Intelligent methods are used for on-demand intelligence delivered through a proactive pursuit of goals and coherent response to stimuli. Agents communicate at a high level of abstraction using an agent communication language. The language provides a universal wrapper for specific content and a capability to negotiate communication means including language, protocol and domain terminology (ontology). Intelligent Agents are easy for on-line deployment and maintenance, and the agent platform running on each device provides an extra layer for engineering network security.