Tip

Four telecom infrastructure trends driving network of the future

Four telecom infrastructure trends -- SDN, NFV, IoT and 5G -- are driving the network of the future, but telecom networks are more likely to transform in 2020, not 2018.

The most publicized technology trends in telecom are software-defined networking, network functions virtualization, software-defined WAN, the internet of things and 5G wireless. A quick review of news coverage suggests these account for more than 90% of all technology news stories written about transformation.

Take SD-WAN out of the mix, because it's not a transformational technology for service providers -- yet. In terms of network operators' telecom infrastructure investment, the other four technologies combined generated less than a 5% change in telecom network infrastructure in 2017, according to data network operators provided to CIMI Corp.

The reason for the disconnect between interest in new technologies and actually deploying them is network operators have sunk enormous investment into their existing infrastructure. To change that, a newer technology like software-defined networking (SDN) or network functions virtualization (NFV) would either need to do things much better at a lower total cost of ownership or produce higher revenue than existing services. Otherwise, it's not worth replacing existing telecom infrastructure. New technologies will help drive the network of the future transformation forward in 2018, but they won't take the place of existing infrastructure until at least 2020.

SDN and NFV developed some momentum in 2017, with examples of deployment for both. SDN is becoming a favored approach for data center switching, and some operators worldwide are talking about adopting white box switches to replace the Ethernet switches they've traditionally bought from major vendors. This trend wasn't strong enough to affect sales much in 2017, and it doesn't get SDN out of the data center.

NFV is of growing interest for deploying virtual customer premises equipment (CPE) functions at the enterprise service edge, which is associated with Carrier Ethernet services. But Carrier Ethernet service opportunities won't bring about a major decrease in network costs or increase in revenue, which will slow operator deployment.

Network of the future needs driving new technologies

What's the difference between these points of early success and the things needed for business or network transformation? The answer is a broader business case. Both SDN and NFV have focused on the limited problem of defining how each technology works at a low level. The bigger questions of how to create broad adoption and generate massive benefits weren't addressed by the standards groups. Although the network of the future will be made up of hundreds of technologies, we don't yet have a holistic vision of how they combine or how they can be made operationally agile and efficient.

Getting SDN and NFV into wider use is going to be difficult, even in 2018 -- and perhaps there's a common reason. SDN is emerging in the cloud data center, and NFV opportunities would be enhanced if operators had more cloud data centers to host virtual functions efficiently. Thus, carrier cloud is the common driver for SDN and NFV. 

SD-WAN could also be essential in driving SDN and NFV progress because it would disconnect virtual private network (VPN) and virtual LAN services from current traditional switch and router technology. While not a transformational technology itself, SD-WAN seems to be on track for considerable growth in 2018. It doesn't rely on SDN or NFV transformation, but SD-WAN facilitates transformation by making VPN services independent of the underlying network infrastructure.

The real question for SDN and NFV adoption is whether there will be enough data center deployments to host more virtual functions and more data center switching opportunities. That depends on an increased number of network operator data centers. If there were more new carrier clouds, which aren't even on service providers' list of top five driving technologies, the SDN opportunity would expand and provide low-cost hosting for virtual functions. But SDN and NFV alone can't drive carrier cloud.

It's easy to see what a network operator could do with carrier cloud, but it's harder to find something compelling enough to justify a new telecom infrastructure buildout. At the technology level, carrier cloud could add more than 100,000 edge data centers worldwide by 2030, according to CIMI Corp. forecasts. This growth would create so many new SDN data center opportunities that using SDN to connect data centers would make sense. This would likely lead to metro-area SDN and widespread SDN success. It would also provide low-cost edge hosting for NFV, which could even make NFV practical for consumer virtual CPE.

IoT and 5G will drive carrier cloud

Without a large initial investment in carrier cloud to jump-start it at a large enough scale, carrier cloud won't be available for general missions in service feature hosting. It won't grow enough in 2018 to make a difference, and that's where internet of things (IoT) and 5G come in.

The internet of things has caught everyone's attention. But so far, telecom providers have had IoT myopia, focusing primarily on the low-hanging fruit of sensor connection. But what companies are spending money to deploy the sensors that telecom providers hope to connect? Whoever they are, they have to earn a good return on their costs. But sensors directly deployed on the internet that are available for over-the-top players to exploit are a pretty hard sell for companies that want to make money on their IoT deployments.

Overall, progress in SD-WAN growth and the 5G FTTN hub mission will decide whether operators see much from transformation in 2018.

Maybe telcos could earn revenue from digesting and analyzing information from sensors to provide any insights. But it's hard to know how regulators would treat that kind of investment. Until telos can be sure of favorable treatment for offering IoT sensor analytics, they don't want to take the risk and make any investments. For these reasons, IoT probably can't help carrier cloud or transformation in 2018.

So, what about 5G wireless? Telecom providers have real interest in 5G. And those that serve major markets face an almost-certain competitive pressure to deploy it. Fifth-generation standards call for a layer of hosted features that can be used to compose services and shared among providers in the same way they currently share connectivity -- via roaming agreements. Specifications for this sort of feature-rich future aren't expected to be complete for 5G until late in 2018, however, and little deployment progress can be expected before 2020.

The big near-term effect for 5G isn't something we hear much about. Many operators have already committed to using the highest-frequency range of 5G -- the millimeter wave -- to create wireless connections from a fiber-to-the-node (FTTN) hub to homes. This telecom infrastructure investment could generate tens of millions of new broadband connections at 50 Mbps or more. Since 5G would likely be associated with streaming video, rather than video channels, it could quickly add to content delivery, caching and ad-insertion opportunities, which are the near-term carrier cloud drivers.

Overall, progress in SD-WAN growth and the 5G FTTN hub mission will decide whether operators see much change in telecom infrastructure transformation in 2018. There's activity in both areas, and while it doesn't look like that activity will be enough to create a total transformation in 2018, we have reason to be hopeful about the technologies used or the network of the future in 2020.

Next Steps

Demand for effective telecom service assurance to increase

Dig Deeper on Network infrastructure