With service providers looking to respond to the fast-growing requirements of consumer and enterprise users, many are turning to network functions virtualization (NFV) as a way to improve network performance, minimize the time needed to launch and maintain services, and lower costs.
The development of NFV brings with it a set of new opportunities – by transitioning to a software-centric network, service providers can bring scalability and elasticity to their operations, improve service provisioning, and offer new revenue-generating services.
According to a report published by SNS Research, "SDN and NFV investments will grow at a compound annual growth rate (CAGR) of about 45 percent between 2017 and 2020, eventually accounting for nearly $22 billion in revenue by the end of 2020."
Clearly, NFV can help address many of the issues service providers face, and will be part of their plans moving forward. However, when making a paradigm shift such as this, the following are a few of the top things to consider to avoid early mistakes when implementing this technology.
- Build out a concrete set of use case with a measurable ROI – As much as CSPs want to believe wholeheartedly in the promised benefits of NFV and dive right in, this is a trap they should avoid. Operators can spend months or years upgrading their network infrastructure thinking they understood the needs of their users, only to realize they missed the mark.
This is why it's important to build out a proper use case that reflects the business value it can add. By carrying out an analytics project to understand the potential uses of a technology, and measure the type of impact it would have on actual incoming orders, then service providers can have strategic plan in place to ensure the success of their investments.
- Implementing key processes from the start – Successful NFV operations depends on the ability to automate. This has been discussed frequently by analysts, vendors, and service providers alike. In order to realize the full benefits of virtualization, automation needs to be implemented from the start.
In the case of telecom, a key aspect to consider when implementing NFV is the concept of automation. Due to its centralized approach to network operations, NFV by it's very nature is reliant on automation. For instance, its possible for a tool monitoring the network to detect the need for more memory in a network function. With NFV, that tool can go ahead and automatically add that memory without involving any human intervention.
- Avoid using legacy processes and systems to monitor the network – Simply put, legacy systems cannot support the business needs of tomorrow's virtual network. Why not? Well, these tools are not flexible or scalable enough to cope with the massive volume of changes and the multiplicity of environments.
While the whole idea of NFV is to build a uniform network, operators can also quickly run into the issue of having a fragmented environment made up of multiple vendors and teams where data is locked in silos. In order to have a clear view of operations, service providers need to have a solution in place that can de-silo, ingests, and correlates all your network big data.
With the enhanced agility, elasticity, and complexity of NFV, a minor infrastructure performance issue monitored by a legacy software solution can rapidly escalate to become a major negative event. To be able to guarantee quality and availability of services, service providers require a next-generation solution that can effectively manage complex, high-value services with the flexibility, performance, and scalability NFV networks require.
CSPs, vendors and industry pundits are all forecasting a significant rise in SDN and NFV projects as virtualization is simply too compelling to be ignored. But in order to ensure the success of their digital transformation, they need to be wary of this mistakes to avoid.