On May 15 the FCC will propose a third round of Net Neutrality rules that will open the door for content providers to pay Internet service providers more money for optimized delivery -- within "commercial reasonableness." Netflix, for example, will be able to pay AT&T for better performance. Think Internet payola.
There's already plenty of protest from those who say a less-neutral Internet will be anti-competitive for upstart content providers. It'll also be more expensive for consumers. On the flip side, service providers want to be better compensated for handling the glut in bandwidth demand.
At this point, it's fair to say we're facing the end of the open Internet. But just how closed the Internet could become will depend as much on technology as it does on FCC policy -- and the court appeals that will surely follow.
Specifically, the outcome here could hinge on the evolution of NFV and SDN. After all, it is these technologies that will allow service providers to dynamically optimize their networks for specific content and services.
Just how closed can the Internet become?
It's important to consider how far service providers will go in charging for differentiated services. Will they charge by content type, device, user location or event? At the Open Networking Foundation Summit this spring, Francois Locoh-Donou, vice president of global products for Ciena Corp., said he was appalled that users couldn't pay for optimized mobile service when 70,000 people descended upon Spain for Mobile World Congress earlier this year.
If consumers can choose varying levels of service for specific applications or content, on what basis will they be allowed to make these choices? If that happens, will we wake up in an environment that resembles the old wireless world where one could never tell what their cell phone bill would look like each month? Beyond that, how degraded will non-premium services become? The FCC says its rules will prevent severe degradation, but how will a decline in service be tested, measured and proven?
Have no fear; legacy networks can't really kill the open Internet
Some of those questions are irrelevant for now. After all, the networks and management systems service providers have today can't enable a super-granular level of Internet fast-laning. In fact, carriers are so intent on developing NFV technology because they want to reach a new level of dynamic service provisioning with granular user access policy. With NFV, operators can provision and string together virtual network functions to support optimized content and combining this with SDN will allow them to do this on virtual networks that stretch across cloud network domains.
More on Network Neutrality
Why NFV and SDN thinkers should know more about Net Neutrality
"If it's as basic as the ISP is deciding that Facebook is going to pay extra so they get priority, there will simply be bigger pipes going into and out of the ISP [which can be done today]. NFV plays no role whatsoever," said Brian Washburn, service director at Current Analysis.
"If we go to a level where the customer gets involved in deciding what content they want delivered at a premium price … that's when you start to talk about different router configurations for each customer -- that's something NFV would need to do," Washburn explained.
Having covered telecom, Internet and cable for many years, Washburn doesn't foresee that kind of granularity happening any time soon. "How long have we been waiting for a la carte cable channels?" he quipped.
However, once it becomes feasible for ISPs and content providers to charge for premium content, it won't be long before consumers will demand the same specialized pay-for-quality experience across their wired and wireless devices -- and that'll also take NFV, said ACG analyst Paul Parker-Johnson.
"We have reasonably well-refined policy decision-making infrastructure in both mobile and fixed services," said Parker-Johnson. Many operators already have sophisticated authentication and subscriber entitlement based on a user's identity and policy type. But being able to "propagate those entitlement profiles among a set of federated policy controllers is really where we have to go further," Parker-Johnson said. NFV will make that possible.
Service providers will also have to enable the same levels of optimized service to users across geographies and operator network domains. A user will want the same premium content delivery through a Verizon cell tower in New York as they can get from a Rogers tower in Toronto. That'll require back-end operator agreements, but the technical side will be addressed using NFV "forwarding graphs," which allows operators to link together virtualized network functions defined by what they need to accomplish. These network functions don't need to be in the same rack or even in the same data center, Parker-Johnson explained.
It'll take three to seven years for NFV to be dependable and implemented at scale -- predictions vary depending on analyst firm. In the meantime, service providers will flex their new found freedom as far as they can with legacy equipment. But while the FCC creates new rules for Internet openness, telecom operators and ISPs will be wasting no time in their labs developing NFV.