ABI's Mavrakis: Live in the Moment for 5G, Cloud Transformation

Only by focusing on the 4G LTE present can telcos master the 5G future, Dimitris Mavrakis, research director at ABI Research, told Telco Transformation in an interview.

In April, ABI Research released its Telco Cloud Framework and Deployment Roadmaps report. In the accompanying announcement, ABI predicted: "large-scale telco cloud deployments will reach global critical mass after 2020, in parallel with the deployment for 5G." The report further states that now through 2022 represents a "pivotal" time for telco cloud deployments -- a timeline that reflects the anticipated development of 5G standards.

To discuss the ongoing implications for the telecom industry of ABI's research and predictions, Telco Transformation reached out to Mavrakis for a Q&A series (lightly edited for clarity and length.) Here, in part one, Mavrakis discusses how the evolution of the market and demands for network packet-core impact and implicate both telco cloud transformation and the network operational transformation from 4G LTE to 5G. He also offers his own advice for telcos and their CIOs and CTOs in gearing up for 5G.

In part two, look for Mavrakis to address the debates, obstacles and risks that surround 5G and cloud transformation.

Telco Transformation: What are the key things for a telco CIO to know about preparing for 5G?

Dimitris Mavrakis: 5G brings many new things to the table. Obviously, technology is just one aspect.

From a CIO perspective, things that 5G will need include a much faster network and also much faster reaction times to the market. So the CIOs -- and the CTOs, as well -- will need to prepare their market systems, and also their core networks, for 5G. And this is why we say that for the telco cloud deployments, 5G will be the catalyst that will really accelerate the need for telco clouds because, simply put, you cannot deploy 5G systems with traditional technology. Well, you can, but it would be too expensive.

TT: So because of the evolution of core network demands, will this lead to a rise in virtualized evolved packet core (vEPC)? What's the vEPC market like presently?

DM: The market presently is not growing rapidly because most 4G LTE deployments have been completed. They are complete so there's not a lot of growth in this market. However, initial 5G networks will be anchored in LTE packet core EPC, This is what we call non-standalone access (NSA) for 5G. So area 5G systems for mobile broadband will be using an LTE packet core, and that will drive some incremental upgrades of EPC.

But after 2020, we expect there will be a lot of deployments of Next-Generation Core (NG Core), which is the fully fledged 5G core network. And in a few years' time after that -- perhaps three, or four, or five years -- we expect NG Core investments to overtake EPC investments. But, on the other hand, there is still a lot of value in LTE. Now, we have gigabit LTE and higher-order LTE categories, so five years perhaps is a moderate time frame for Next-Generation Core overtaking EPC, but you never know.

TT: Since ABI Research's Telco Cloud Framework and Deployment Roadmaps report was released in April, has anything changed in the market?

DM: No. Not much. Because the challenge of implementing telco cloud is not the technology; it is the operational or the cultural inertia of telcos and of service providers -- and that cannot change within six months, as you know, or even one year or two years. And that's why most NG Core deployments will be greenfield. They will start from scratch, where telcos start to implement new business models.

TT: For that reason, might telcos in developing nations have a leg up here in 5G transformation because they are less attached to a legacy model?

DM: Probably not because it's almost certain that Next-Generation Core is going to be more expensive than EPC. In a developing market, especially in a market that is challenged in terms of profitability, it's not likely that telcos will start with 5G. It's more likely with 4G, or even in some cases, with 3G as well.

But what is interesting is Tier-3 telcos -- the smaller telcos, not the incumbents or the number ones or number twos. The smaller ones are more agile, but that is a much longer discussion because with telco clouds and 5G, there is a level of self-integration that is necessary. A telco like AT&T needs to integrate some systems themselves in order to maintain some level of independence from vendors. And this is where the Tier-1s have an advantage. But that's not saying that this model will be the most successful deployment model. It may be that the Tier-2s and the Tier-3 telcos will partner with vendors to grow faster with Next-Generation Core. So to summarize, it is too early to be able to say how exactly this will develop.

By the way, by January we will have what is called the early drop for the first 5G standards for 5G New Radio in non-standalone access, which means that it's a new radio network with LTE packet core. The full standard will come in June next year. And for Release 16, which includes Next-Generation Core, the core network architecture, the protocols will come December 2019 or early 2020. (See: 3GPP to Unveil First 5G spec.) Really, we cannot expect to know anything before that about 5G core network. At the beginning of 2020, we can start discussing deployment issues of Next-Generation Core. Everything before that is conjecture.

TT: How is the race to virtualize impacting impending 5G deployment? And vice versa?

DM: At the moment it isn't because early 5G systems will be mobile broadband with a 4G packet core with an EPC. So it isn't; it isn't affecting it directly. Perhaps it will affect it indirectly because with 5G you need much more capacity -- so you need VNFs, NFV and SDN. That does not mean you cannot do 5G early without a telco cloud deployment, but, in two or three years' time, when we have Next-Generation Core, then it becomes a necessity. So SDN and NFV become the key building blocks of 5G Next-Generation Core.

TT: So it kind of sounds like "First things first" -- that rather than focus on 5G itself, telcos should focus on what's directly in front of them -- which will better prepare them for 5G. Does that sound right?

DM: That's right. That's exactly what we say. Telcos need to try to work with what they have today and try to change their operations to be more agile and more open -- to use open source and so on. And then they will be prepared for 5G before 5G. So, an example: 5G is all about verticals -- so automotive, augmented reality (AR), virtual reality (VR). Telcos have no clue about these markets yet; they don't know how to sell to those markets. So what we say is that telcos need to start thinking about them, and perhaps start experimenting within these markets even with 4G. And this will give them the experience they need. When 5G comes and is deployed, then they will be much more equipped.

— Joe Stanganelli, Contributing Writer, Telco Transformation