Automated service provisioning in heterogeneous large-scale environment

With the increasing complexity of network management activities due to naturally limited human involvement, carriers and service providers are looking to migrate from manual, static provisioning models to the more dynamic service-oriented automated provisioning models to meet customer demands for rapid service turn-up and obtain more customers and maximize revenue opportunities. We propose a novel distributed architecture where highly mobile and intelligent agents can take the responsibilities of not only provisioning, but also configuration audit management in a timely fashion. Although previous research has focused on using mobile agents for network monitoring or simple push-based device configuration in a distributed architecture, their ability have not been exploited in dynamic IP service provisioning. Using a simple push-based model to configure network services while ignoring dependencies among configuration elements may easily lead to configuration inconsistencies resulting in failure or inefficiencies. In this paper, we have taken a new approach to configuration modeling that is device neutral and based on which any existing or emerging IP services can be presented by encapsulating service semantics, including service-specific data. We have developed new mobile intelligent provisioning and audit agent architectures that use the knowledge built upon configuration dependency modeling. Examples of intelligent agents are presented to complement the proposed management architecture.