In The Spotlight: Is NFV a Solution Looking for a Problem?

Author: Kim Gibbons, Chief Marketing Officer at NetNumber

NetNumber LogoAt this month’s LTE World Summit 2014 in Amsterdam, one question seemed to run thematically through several presentations during the Signaling Focus Day – is NFV a solution looking for a problem?

The desire for OPEX / CAPEX savings is driving carriers to consider implementing NFV and SDN approaches offered by many vendors today.  Given the increasing focus on NFV from vendors, press and analysts, and event management firms – the question is reasonable.

In our viewpoint, NFV provides value today where the hardware footprint is oversized for the service provided. Signaling can benefit through the use of COTS hardware as a first step into a flexible and dynamic world.

We see many vendors jumping on the NFV bandwagon by replicating their physical hardware and software layouts into virtual environments.  This approach was evident in several vendor presentations during the LTE World Summit.  Unfortunately, such an approach does not deliver on the promise of NFV – improving efficiencies, increasing agility, and lowering costs.

If each function is still separate when virtualized, it must be deployed, managed and updated separately.  Virtualized chaos is still chaos!  This approach does nothing to reduce complexity in the network core.  And complexity is a major barrier for carriers that want to change or offer new services quickly.

NetNumber has been virtualizing network functions for 14 years, long before it was the buzzword of the day.  Our approach is to provide a software-based, common infrastructure for the virtual delivery of real-time signaling and control, combined with database services. This platform is easily scaled and distributed to the optimum spots in a carrier network.  The result is a radically simplified network core.

Perhaps the question industry should be asking is not whether NFV is a solution looking for a problem, but rather, how can we help carriers embrace software-based solutions as key to evolving their core network infrastructure.



You may also like...

1 Response

  1. I agree with the author, but the issue may be far deeper and more complex.

    Service providers are really just edge access providers to “core content, application, and service providers”. They have a narrower view of demand than the edge providers.

    Since they do not have a complete view of edge demand they cannot scale and price properly ex ante.

    The result is cost-based pricing (based on relatively high average costs because of redundant and inefficient capex and opex at every layer due to vertical integration mindsets) as opposed to value-based pricing (based on marginal cost that clears marginal supply and demand cleanly.)

    But at the same time, since the IP stack lacks settlement systems with its bill and keep approach, there are no effective price signals or incentives to clear marginal supply and demand north-south (between app/content and infrastructure) and east-west (between network/agent A and B) in the informational stack.

    We need to move from a supply-centric orientation and debate with regards to NFV to a demand-centric approach which includes pricing and demand elasticity.

    Michael Elling

Leave a Reply

Your email address will not be published. Required fields are marked *