Mallinson: Net neutrality could cut off new use cases in a 5G world

Keith Mallinson

Net neutrality rules resulting from the FCC's decision to reclassify broadband, including cellular communications, as a telecommunications service under Title II of the Telecommunications Act in the U.S. could undermine what is being proposed for 5G networks by the Next Generation Mobile Network (NGMN) Alliance and others.

President Barack Obama demanded that FCC ensure all Internet traffic be treated equal in its new net neutrality rules; but different types of traffic have markedly differently network performance requirements and make vastly differing demands on network resources. Restrictions on offering a variety of service profiles and pricing them accordingly will hobble 5G's enormous potential. Imposing such Soviet-style uniformity on Internet traffic would be a travesty.

There are compelling engineering and economic reasons to unify and virtualize network capabilities and make Internet-based a wide range of services including voice calling, messaging, information access, file transfer and the streaming of entertainment content. However, proposals for 5G clearly illustrate that different use cases have widely-different network performance requirements. These also significantly affect network design, resource allocation and costs.

Five years ago, I expressed my concerns here about the dysfunctionality of imposing net neutrality conditions on mobile network operators in particular. I continue to believe that such restrictions will be harmful economically and impede innovation which would beneficial to consumers. Some safeguards are necessary to ensure vertically-integrated operators who are dominant in networks do not unfairly discriminate against their over-the-top competitors downstream in value-added services such as messaging, information and entertainment delivery. For example, transparency is essential. Class-of-service parameters and quality-of-service actually achieved should be made clear to customers. However, mobile operators are rarely dominant in the way that fixed network telcos and cable TV companies often still are because there are almost invariably multiple mobile operators and significant market share fragmentation.

One size does not fit all

On Nov. 10, 2014 "President Obama call[ed] on the FCC to take up the strongest possible rules to protect net neutrality, the principle that says Internet service providers (ISPs) should treat all internet traffic equally." Amusing remarks by talk-show host John Oliver, among other commentators, spurred the crowds to demand this populist outcome by submitting comments to the FCC. Who would want to pay extra to stream 4K Netflix? On Feb. 26, 2015 the White House reported that "the FCC just voted in favor of a strong net neutrality rule to keep the Internet open and free. That happened because millions of Americans across the country didn't just care about this issue: You stood up and made your voices heard, whether by adding your names to petitions, submitting public comments, or talking with the people you know about why this matters."

But treating all data packets the same is network engineering and economic nonsense. We transport coal, people and first class letters differently for good reasons. Similar logic should apply for widely different types of telecommunications traffic.

In the old days and still, but to much a lesser extent now, different telecom use cases are served by distinctly different networks. Mission-critical corporate networks were strung together with high-availability leased lines while public packet-switching networks were "best effort," managed data services on top of telco transport offered some service-quality assurances, telex was slow and limited to basic text but provided legally-recognized message delivery, the PSTN delivered high-availability voice, and mobile service levels were much lower due to network coverage and capacity constraints. The Internet has also been largely "best effort" in its core, while accessed with low-bandwidth, high availability dial-up, or higher-bandwidth, high-availability leased circuits, or best-effort connections with DSL or cable modems.

One network optimized for many and widely different use cases

The scope of the Internet has expanded massively to engulf most telecoms-based services: from email, file transfer and the web to include social networking, phone calls and real-time streaming of audio and video, including substitution for broadcast consumption. OTT providers including Google, Skype, Facebook and Netflix are significantly responsible for growth in all of these categories. Internet access has also expanded very significantly, with usage becoming predominantly mobile-driven, via smartphones in particular. It seems that pretty much everything in telecom for consumers, including mobile, is now Internet-based, or at least there is a broadband or Internet-based alternative to telco or mobile operator service, including, for example, Skype and FaceTime for voice and video calling, and WhatsApp for texting.

Network infrastructure development, with prospects for 5G mobile, was back in the limelight at the Mobile World Congress in Barcelona earlier this month. Devices, including smartphones in particular, have grabbed most attention in recent years. Among conference sessions on 5G, NGMN held a press conference to announce the finalization and publication of its 5G White Paper. It introduced this by stating that "[a] global initiative has delivered key end-to-end operator requirements intended to guide the development of future technology platforms and related standards, create new business opportunities and satisfy future end-user needs."

According to NGMN, "[i]n addition to supporting the evolution of the established prominent mobile broadband use cases, 5G will support countless emerging use cases with a high variety of applications and variability of their performance attributes: From delay-sensitive video applications to ultra-low latency, from high speed entertainment applications in a vehicle to mobility on demand for connected objects, and from best effort applications to reliable and ultra-reliable ones such as health and safety. Furthermore, use cases will be delivered across a wide range of devices (e.g., smartphone, wearable, MTC) and across a fully heterogeneous environment. NGMN has developed 25 use cases for 5G, as representative examples, that are grouped into eight use case families. The use cases and use case families serve as an input for stipulating requirements and defining the building blocks of the 5G architecture. The use cases are not meant to be exhaustive, but rather as a tool to ensure that the level of flexibility required in 5G is well captured. The following diagram shows the eight use case families with one example use case given for each family":                                            
5G use case families and related examples (source: NGMN 5G White Paper)

5G is conceived to be a unified network which exploits flexibilities, efficiencies and economies in virtualization and software-defined capabilities while delivering a wide variety of service profiles. Different use cases have widely differing performance requirements. Criteria include network speed, mobility speed, latency, service levels and availability, power consumption and terminal density (i.e., per square-kilometer). In fact, the ability to achieve some performance criteria is in direct conflict with others. For example, the fastest access speeds, highest mobilities, lowest latencies and highest reliabilities are not consistent with longest battery lives or lowest costs.

Some discrimination is good

In rebuttal to such unqualified demands to "treat all internet traffic equally" by President Obama and others, providing different service performance specifications and classes of service among different use cases and users can benefit everybody so long as this is done even-handedly. For example, it could be done between direct competitors in the same market segment. Networks are capital intensive and subject to significant economies of scale. However, differentiating services according to use case requirements and pricing according to demand elasticity and with respect to when capacity is most plentiful or scarce (e.g., on basis of daily demand cycles) will help maximize network utility, customer satisfaction, economic efficiency and investment incentives.

These net neutrality rules will be counterproductive to the objective of keeping the Internet open to innovation and competition. Regulation with Title II in such an emergent phase will be difficult and could be very harmful. Innovative and popular offerings such as free music streaming with T-Mobile US' "Music Freedom" have been identified as prospective violators of net neutrality rules. This is contested by T-Mobile CEO, John Legere, but it is far from certain he or others who seek to introduce similar offerings will prevail. It would be a crying shame to stop such demand-stimulating innovations in their tracks. And who knows how very well 5G might enable development of other innovative new services, business models and pricing packages if it remains unshackled?

Keith Mallinson is a leading industry expert, analyst and consultant. Solving business problems in wireless and mobile communications, he founded consulting firm WiseHarbor in 2007. Find WiseHarbor on Twitter @WiseHarbor.