Using multiple collaborative agents for adaptive Quality of Service management of C4ISR networks

摘要

This research explores the potential of agent technology for adaptive Quality of Service (QoS) management of C4ISR networks. With the growing emphasis on information superiority, any time savings or additional utilization of resources enabled by effective network management becomes increasingly important. Intelligent agents are ideal for assessing information, adapting to dynamic conditions, and predicting future network conditions. In the kernel of the proposed multiple agent system (MAS) testbed are agent shared memory and majority rule architectures for agent conflict resolution. The case based reasoning (CBR) technique provides the foundation for building the agents' shared memory of QoS management solutions and allows the individual agents to share their associations of feedback controls in response to application and user QoS profiles. Based on the Telecommunications Management Network (TMN) functionality, we use this agent architecture to effectively translate the warfighter's service layer application requirements across the network. The fundamental frameworks of Service Level Management (SLM) and Policy Based Management (PBM) serve as cornerstones in effectively gathering and applying specific application requirements. Finally, we utilize these techniques to investigate an actual C4I application at the Pacific Region Network Operating Center (PRNOC) in Wahiawa, Hawaii as the real-world focal point of the thesis.