Operations

As NFV trials proceed and service providers begin to understand the intricacies of building hybrid virtual and physical networks, the next challenge is how to operate them. From fulfillment platforms to service assurance to customer care, there are a lot of details to work out in the operations arena. While network engineers might believe their job is done when the NFV elements are integrated into the network, the challenge for operations is just beginning.

Beyond cost savings, most service providers hope to use virtualization to bring new products to market more quickly. The challenge then is not the network, but the OSS/BSS platforms used for product development, service launch, fulfillment, and assurance. Existing fulfillment and assurance solutions are cumbersome, clumsily integrated, and often require manual efforts to complete. The volume and variety of new services intended to be enabled by NFV simply cannot be delivered using today’s maze of OSS/BSS implementations.

Early adopters are implementing an orchestration layer over existing network infrastructure to create a hybrid network. To do that, service providers must develop numerous application program interfaces (APIs) and tons of customized integration to enable automation of fulfillment and assurance functions. The orchestration layer is fully automated and greatly improves delivery time for new services while eliminating error-prone manual tasks. However, all those APIs and integrations have to be meticulously maintained and regression tested every time there is a change to real or virtual network elements.

Start Smart

In the pursuit of NFV, it is important to recognize that no single virtualization strategy will be optimal for all network functions. A hybrid network strategy bridges appliance-based networks to NFV and implements an architecture optimized for performance, reliability, cost, and customer experience. To ensure interoperability and consistent implementation of virtualization in the public network, service providers will require standardization of APIs, management data, and control. A fully-automated and standardized orchestration layer is the key to delivering both physical and virtual network functions economically and at scale.   

There are valid engineering reasons to implement network functionality in both hardware and software. Time-to-market for new products and reduced costs are arguments for NFV; while speed, performance, reliability, and visibility are among the reasons that an appliance may be preferable to an application. Not every piece of hardware will be replaced and not every piece of software will perform as it should, so it is important to understand the entirety of each function and how virtualization will affect the network, services, OSS/BSS, and customers both now and in the future.

Nancee Ruzicka is the principal analyst at ICT Intuition and a contributing editor to Pipeline.