No Jitter is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

The "Internet Fast Lane": For More Than Video Buses?

It’s pretty clear now that we’re entering a new age of Net Neutrality where the FCC’s position is much more focused, even forgiving.

While the new rules aren't published yet, FCC Chairman Tom Wheeler has made it pretty clear that these new rules will not prohibit provider-pays models of QoS or settlement for QoS among ISPs. That could certainly mean the potential for higher-quality video (at a price) but it could also mean that higher-QoS Internet service could become available for other kinds of traffic, and that, in particular, businesses might be able to use the Internet for things they now have to use business VPN services to support. What might the "fast lane" do for business users?

One obvious benefit would be for video collaboration or simple video meetings. Today, high-quality video requires private networking facilities that are so expensive that for anything but regular use, costs would be prohibitive. Since most video use is extemporaneous, a video-QoS model for the Internet could revolutionize video collaboration.

This revolution might be at the expense of not only the providers of business VPN services, but the providers of high-end videoconference tools. Even today, some surveys show that Internet videoconferencing generates ten times or more the number of conferencing hours as high-end systems, even if you factor in the larger number of participants per room in the latter.

The notion of extemporaneous QoS for video could also spread to other business applications. How much private networking is justified by the higher availability and guarantees of service quality that such a network service brings? Would companies who could invoke service-specific QoS on demand decide to try even critical applications over Internet VPNs, with the intention of dialing up QoS if they needed to? How much best-efforts-suitable stuff gets carried on expensive VPNs because a small percentage of companion work demands high QoS and availability?

Cloud computing could obviously benefit from QoS and settlement, too. Many companies who are considering public cloud services want service-level agreements. But if you get a cloud service over an Internet connection that's best-efforts, what does your SLA mean to you? Not much. On the other hand, pair cloud SLAs with Internet QoS and you have something that has contractual guarantees of the kind that make a CFO (almost) smile.

Then imagine what might happen to the business opportunity for a small cloud provider. Today, they might be selling cloud door-to-door in some small city somewhere because they can't get a service footprint any bigger. Now they could become an ISP, get QoS peering with other ISPs, and sell cloud business and even VPN services anywhere in the world, right alongside the giants of networking who have multinational footprints. Think innovation, David and Goliath, or whatever you like here--it would be a radical change in the business model.

There could be ISP changes coming, too. Settlement could make retail ISPs profitable again. Right now a startup trying to get VC funding for fiber to the home (FTTH) would be better off starting a dating service for left-handed lacrosse stars; neither idea would have much revenue credibility but at least the DSLHLS service would be cheap to fund.

But now suppose that content providers and other ISPs who want quality video or other services delivered to consumers have to revenue-share by paying that retail ISP? That might re-ignite the FTTH business.

This happy future isn't a slam dunk, though. For extemporaneous QoS to work, we have to be able to get it wherever we get the Internet, and that doesn't mean pay-for-priority so much as it means settlement among ISPs for the payment. Ad hoc video or application use of QoS is almost certain to cross ISP boundaries. If you pay one for priority, and the rest just toss your stuff into the bit bucket, nothing has changed. While the FCC may permit QoS and payment for it, it may fall way short of mandating settlement.

Why? The answer is the D.C. Court of Appeals decision that got us where we are with neutrality. The court said that the FCC has only limited jurisdiction over ISPs as long as it continues to say that the Internet is an "information service." That could mean that an attempt by the FCC to mandate that an ISP offer QoS services for all applications or settle with partners for end-to-end QoS would be overturned by the court for the same reason.

"Real" core net neutrality means non-discrimination when managing traffic. That's a given. The FCC is signaling that it also means paid-priority options regardless of whether the consumer or provider of content pays. They're saying that this kind of fast-lane service can't be allowed to favor a provider's subsidiary over others, which implies the FCC would say that if you have a fast-lane service you have to open it to whoever pays the toll. But it doesn't mean that you have to offer fast-lane service at all. Remember, consumer-pays-for-priority has always been an option, and nobody offered it.

Still, I don't think the ISPs would be able to resist the opportunity to charge video companies for priority delivery, and that's the camel's nose here. From that we can see a pretty clear path all the way to everyone-must-offer-priority. The hang-up will be settlement. If the FCC can somehow navigate the fine line the court set on its jurisdiction over "information service" providers and mandate QoS settlement, then we may see a radical new Internet--and network--on the horizon.

Recommended Reading: