Just when you thought your network was getting pretty good, it seems the new marketing benchmark in 2017 will be gigabit speeds. Not just for fiber deployments but for every access medium, spanning copper, cable, and wireless operators – and that’s a first. But there’s more than just that coming. Here’s a run-down of what we’ll see in 2017:
Copper deployments will meet the speed challenge using G.fast, a technology that utilizes the noise cancelling technique from VDSL called “vectoring,” in addition to enhanced modulation schemes to get more and more out of what’s in the ground today. In wireless, increased levels of carrier aggregation and the usage of unlicensed spectrum (LAA) will help do the same in LTE, while 5G trials will go beyond a single Gbps. Optical has, and always will have, a physical advantage due to a pristine fiber medium, and here we imagine 10 Gbps (XGS-PON) starting to become the de facto planning base for new deployments – with an eye towards 25 Gbps and even 100 Gbps in the future. Not to be outdone, we also see hybrid fixed and wireless solutions coming to the market to help deliver gains not possible with a single technology.
All of this speed will not come without architectural changes to operator networks. In both the fixed and wireless domains, speed will come at the expense of shorter loop lengths and cell sizes. This shrinkage of service area will require fiber to backhaul all that traffic – closer and closer to the edge. Likewise, the demand for millisecond latency targets implies some services will also move outwards, thus creating “edge clouds.” Edge clouds will be small scale central office like deployments that will reduce the need to hairpin traffic through a few large centralized sites, and provide unique localized services. Edge clouds. Remember that.
Distribution comes at a cost. It implies significant optical and IP investments (which we like!), but also more attention to self-configuration, self-operating, and self-healing networks. The reason lies in the inherent complexity that comes with distribution; a complexity slowly reaching a stage where it is beyond human capability to manage. And thus, automation becomes key. To that end, SDN will become mainstream in 2017, along with increased investment in SON and analytics, using machine learning to give the end customer an increasingly personalized and more powerful network experience, while keeping costs affordable. These concepts are not new, but distributed networks will move them from being niche increments to operations to fundamental baselines.
5G: From PowerPoint to Metal
Nowhere will distributed networks and edge clouds be more prevalent than in 5G. In 2016, 5G largely equated to great promises and Proof of Concept systems. More cynically, great PowerPoint decks. In 2017, reality will start to set in as trial systems start being deployed based on published specifications. It is from these trials that the truth on the practical implications of millimeter wave deployments, viability of the use cases being considered, and the cost forecasts of e2e solutions will become self-evident. As we near the end of 2017, we imagine a more fact based view of 5G, enabling better forecasts of commercialization timelines and the business potential ahead of us.
Virtualization of Everything
With all this talk of the future, we shouldn’t forget legacy systems. Aging technologies like wireline PSTN, bare metal cable solutions, CDMA, WiMax, and GSM are due for obsolescence, albeit in good time, as millions of users continue to depend on these services. But death of the aged, usually means a path to enlightenment and an opportunity for new and innovative solutions. And that’s what we see. Virtualization will continue its unending path towards being the new norm for wireless in 2017. However, we now see it being applied to the existing fixed networks such as the HFC networks to reduce power and space constraints at the hub office.
Really Great Video
What services will take advantage of this speed and capacity? Unicast video and virtual reality are, without a doubt, at the cusp of large scale adoption. As nearly every show that exists becomes available through streaming, and nearly every broadcaster offers a streaming option, it’s clear unicast video is on a classic hockey stick trajectory. Likewise, with Google, Facebook, and Samsung offering commercial hardware VR systems, we see VR and AR gain interest in 2017, with only better hardware in 2018 and 2019 holding back large scale adoption.
For IoT, 2017 will finally see the launch of competitive cellular solutions with Cat-M and NB-IoT deployed network-wide across most carriers. The open question then is what happens with the historically lower cost alternatives such as Lora, Sigfox, Z-Wave and Zigbee? For some time to come, we imagine a happy co-existence of all, each serving an appropriate niche, albeit with low cost cellular IoT growing in market share. The question that remains though, is how much share?
2017 will be an exciting year for new launches, with Gigabit access, VR, 5G trials, and cellular IoT. Those services will require architectural changes and will increase complexity to networks, however, mitigated by increasing automation with SDN, CEM, analytics, and SON. In short, it will be a thrilling year, ushering in changes we’ve been talking about and readying for some time. The fun part will come in 2018 when we see how it all went, but that’s for another article.
Mike Murphy is Nokia’s CTO for the North America region. David Eckard is Nokia’s CTO for Fixed Networks. Nutan Behki is Vice President of Nokia’s Applications & Analytics business.
Filed Under: Infrastructure, IoT • IIoT • internet of things • Industry 4.0, Virtual reality