|
|||
|
OPNET
Technologies OPNET is a
registered |
1. Autonomic Network Management As demand for data services grows rapidly, service providers are faced with the challenge of how to bring their cost structure down, so they can generate sustainable margins on competitive services that are typically priced lower than the traditional private line services. Lowering the cost structure of data services naturally involves a reduction in the capital expenditure, but even more importantly, it necessitates a revolution in the operational expenses of provisioning, maintaining and monitoring these services. Many carriers today operate multiple networks, each of them responsible for transporting and switching just a single type of traffic for a single service. Since operational expenditure typically accounts for the majority of a carrier's cost structure, it is not surprising that "Convergence" has been an important goal of most Service Providers. Consequently, one of the main themes of the next generation network is the ability to efficiently and flexibly multiplex and aggregate multiple services over the same physical facility. This ability enables a provider to provision new services on a pre-provisioned facility in a way that doesn't require an upgrade of the physical plant with every new service turn-up. Multiplexing multiple data streams, including those that carry different layer-2 protocols require some kind of a generic tagging mechanism to distinguish between different flows. MPLS is the only technology available today that can provide generic tagging of multiple types of data flows. The control and management of large-scale network centric distributed applications over a heterogeneous and dynamic execution environment presents challenging research problems. A possible solution to these research problems inspired by biological systems (autonomic nervous systems), which are very successful in controlling and managing complex, interactive, and constrained systems. We refer to this approach by Autonomic Computing. Autonomic computing involves computing systems and services that are capable of self-configuring, self-organizing, self-healing and self-protecting; and as a result they can anticipate their workloads and adapt their resources to optimize their performance. In this project, we address the service provider operation expenditure issue by designing an autonomic system to control, manage and consolidate the Core/Metro network resources in a way to be able to deliver IP services at specified service levels based on a pre-defined SLA. We assume an IP/MPLS technology for the service providers' core network based on our previous argument. The proposed Autonomic Architecture will monitor, plan, analyze and execute the network actions so that the overall system looks pretty stable and autonomous.
|
