The original Perfect Storm was a weather event that happened between October 28th and November 4th, 1991. The book and movie traced the Andrea Gail fishing boat and its crew, from Gloucester, Massachusetts. Now, 30 years later, we are entering an equally turbulent time; a perfect storm of a different kind.
5G, on its own, represents a significant disruption for carriers. However, it comes at the same time as other technology and architectural options become available as well; each disruptive in their own right. Those include continued capacity growth, use case expansion, vRAN, O-RAN, network distribution and webscale participation in telecom networks. It is this confluence that creates The Telecom Perfect Storm. The challenge is in deciding which are appropriate at a given time and in which situations. The answers are not straight-forward.
Perhaps the easiest starting point when looking into such an abyss is to extrapolate. In that regard, for capacity, or Petabytes of data, while there has been some reduction in demand in urban centers due to COVID, we expect growth to return and eventually exceed the ability of networks to service the demand. In short, LTE will run out of gas, or hit exhaust at some point. If allowed to occur, that would manifest itself as a significant deterioration in user experience. In the study shown below for a major metropolitan area in the US, that starts around 2022-2023.
The only way to prevent exhaust is to deploy 5G in new frequency bands, with gusto, such as C-Band and/or mmWave - we do not believe WIFI or a reduction in demand will happen as some suggest. The approach used in this particular Nokia simulation can also identify indoor vs outdoor requirements and thus, in total, addresses the where and when for 5G deployments.
Mike Murphy,
CTO, Nokia
Americas
Unfortunately, capacity demand is perhaps is one of the more predictable areas of telecom. The other disruptive elements noted previously, like storms themselves, are highly unpredictable in timing and impact, making decisions about their adoption more complex. But let’s try.
Use case expansion is a complex topic and requires perhaps more space than afforded in this article and might be a good follow-up! However, in the interest of completeness, perhaps what we can say for now is that basic needs, such as mobility coverage and capacity, need to be serviced first, because they are enabler for the true promise of 5G, i.e., subsequent use case expansion.
Open RAN solutions have been synonymized with the O-RAN set of specifications of which there are many, at varying degrees of maturity. The complexity of the fronthaul specifications, perhaps the most important, and where Nokia is a key contributor, has forced the creation of profiles that limit the number of options required to enable realistic plug and play implementations by vendors. 106 profiles have been defined to date and the numbers are growing. Operators and vendors alike, are just now starting to align on those considered the most appropriate for specific deployments.
Independent of profiles, is the issue of feature parity. Solutions deployed today by new vendors do not support feature parity with those deployed by the large incumbent vendors. For some situations, like rural, private or greenfield networks, that might be ok. However, for most Tier 1 carriers serving Tier 1 markets, it likely is not. In short, we see O-RAN compliant solutions as requiring more time to be ready for large scale global deployments.
Then there is vRAN. While vRAN and O-RAN have often been conflated, they are technically orthogonal. To date, virtualized RAN solutions (in particular, DUs) tend to have higher CAPEX than bespoke solutions. This is reducing over time, but is not there yet. That drives justification of DU virtualization based on operational gains. Those gains often require significant changes in how carriers buy, deploy, architect and manage their networks and is further complicated by the need for a duality of network solutions for a long time to come. For example, we often think of virtualization driving some degree of centralization in CRAN Hubs. Hubs require real estate and fiber investment. Few carriers have done this, and until others have, large scale, global, fully virtualized RAN deployments will also be muted.
Finally, in 2020 it was hard for a week to pass by without some announcement of cooperative plays between carriers and webscale or hyperscale companies. This was partly driven by webscale interest in expanding their opportunities for load volume growth and 5G applications, and carriers realizing internally developed platforms were difficult to maintain without greater scale. This is perhaps the one major evolution that is indeed ready for prime time, at least for core network functions and OSS/BSS systems and in the not too distant future, RAN functions will be added as well. However, the definition of edge differs between webscale and carrier, as do cost expectations. So, while ready for prime time, it is perhaps also astute to proceed with caution.
Given all that, what does one do in a Perfect Storm when there are so many shiny objects ahead with mermaids trying to draw one’s attention into the deep? There are likely more opinions on this than the number of options in O-RAN profiles! Being prophetic across all possible scenarios is impossible. However, when we think of large-scale deployments by Tier 1 operators in Tier 1 markets, the outlook is slightly more predictable. If we start with the knowns, that means rapid deployment of traditional 5G systems to meet urgent capacity demands, especially in 2021 That could include broader webscale adoption as a platform for certain functions. Once that baseline has been established, expansion into new use cases can follow. And while vRAN and ORAN need some maturity for large scale global adoption, some degree of experimentation through PoCs or trials seems prudent, with expansion based on individual carrier and ecosystem readiness.
It’s a tough time to be a captain or CTO during a Perfect Storm. But it’s also a time to make decisions that can drive incredible operational improvements or revenue growth. And that is indeed an exciting journey.
