Like other technology generations before it, 5G is at an awkward stage. With initial standards specification well under way, the happy days of bluesky thinking and reality-free speculation are over. But until the 5G New Radio is deployed in anger, it is too early to draw very sound conclusions about how it will perform, how it will impact operators’ business models, or how quickly it will be adopted.
It is possible, now, to make some forecasts for the early years of 5G. Rethink Technology Research added 5G to its forecast of mobile base station and site deployments in the first quarter of this year. We predict that, assuming there are some pre-standard roll-outs in 2018 and standards-based deployments from later in 2019, operators will build out about 440,000 5G macro and micro base stations by 2022. About 60% of these will be flexible multimode (4G/5G) units, while about 15% will support a pre-standard implementation such as Verizon’s. A large proportion will be incorporated into virtualized architectures and will be densified over time with an accompanying layer of small cells.
It is only from 2022, however, that 5G will really be starting to be a mainstream option – for large numbers of MNOs deploying networks over large areas of coverage, rather than specialized hotzones for venues or city centers. Until 2021, there will be a flurry of activity by the small group of cutting edge players, and from small-scale deployments by others, but that will almost certainly be accompanied by something of a backlash as business models, key use cases and cost structures take longer than anticipated to solidify.
Analysts at SNS Research have come up with their own forecast, this time on the revenue front, predicting that operators will spend more than $21bn on standardized 5G NR infrastructure by the end of 2025, plus $7bn on new next generation core and transport infrastructure. They point to the GSA’s statistic that 25 operators in 15 countries have now demonstrated pre-standard 5G technologies, or announced trials. But it still seems likely that, if their forecast proves sound, most of those dollars will be spent towards the end of the period (and we suspect the ratio of RAN to core/transport will be far lower).
For now, operators are engaged in two inevitable preambles to a full standard – boasting about early trials, and taking sides in a standards split. The split is between Verizon’s own specifications, which have leapt ahead of the 3GPP but which will require some effort to bring into line with finalized standards; and 3GPP’s own 5G NR Non-Standalone. This is a subset of the full 5G NR, which still requires an LTE anchor network. By extracting this portion of the standard, 3GPP was able to agree to a request by a group of companies, led by AT&T, to accelerate its publication schedule. The initial spec will be finished by the end of this year and expected to be approved by March 2018, allowing for standards-based products and deployments in early 2019.
Opinion is divided over the merits of each of these fast-track approaches – or indeed, whether there is any real need to move to 5G so quickly anyway. European MNOs have been particularly negative about the arms race which is building up, even if this will be more short-lived than 4G’s WiMAX versus LTE, or the GSM versus CDMA, debates, since all the parties are committed to convergence as soon as possible.
Enrique Blanco, the influential group CTO of Telefonica, told Mobile Europe recently that speeding up the 5G standardization process risked stripping out many of its benefits, and resulting in a platform with limited differentiation from 4G.
“My personal view is those operators who are really pushing for this – in the US, and Japan, and Korea – have different, more urgent priorities,” he said. “Because of the Olympics, and whatever else, they want to deploy services using NR much sooner – which means they need the standards to be defined sooner, and the chipset makers to deliver sooner.” But he added: “5G is not just about sub-millisecond latency and 1Gbps speeds – we can get close to that with 4G already. So what’s new? The business model needs to be defined; 5G shouldn’t just bring new radios and antenna systems but new network architecture.”
The full interview with Blanco can be read in the forthcoming April/May issue of Mobile Europe – see here.
Blanco has spoken before of the real importance of 5G lying in fundamental architecture changes, enabled by virtualization and software-defined networking, which will, in turn support new business models based on slicing and full agility. But none of this transformative capability will happen with non-standalone 5G NR, and he thinks 5G roll-out will take place over a longer time period than many expect, not a shorter one, because many of its key enablers, such as Massive MIMO, are being deployed first in existing 4G networks.
Orange’s director of technical strategy, Yves Bellego, also thinks focusing on a limited range of radio specs carries high risk. Commenting on the March agreement to a two-speed 5G NR, he said in an interview: “5G is not and should not be limited to NR. We really believe we need a full 5G system… and there is a risk that defining a new radio and not addressing other portions does not in the end deliver the best solution. Having new radio does not answer all the needs.”
Nokia has also been surprisingly cautious, for a vendor, about the acceleration. It did not sign up to the AT&T-inspired 3GPP initiative, perhaps because it has been focusing even more heavily than its rivals on packet core issues, and on opportunities enabled by advanced virtualization, mobile edge computing and slicing. In its official statements, Nokia just said it did not want any distraction from its 5G efforts and so had not joined the acceleration group. And it is keen that the RAN standards are developed in tandem with those for the core and systems architecture, not before them, or risk a disconnect which will compromise new business models based on approaches like slicing.
None of this reticence has stopped both fast-track groups moving ahead. Last week, Samsung, Cisco and Verizon announced the first multivendor, end-to-end trial of the US operator’s 5G specs, in the Detroit suburb of Ann Arbor. This will be followed by fixed wireless pilots in 11 US cities before midyear.
In the Ann Arbor test, Cisco provided a virtualized packet core from its Ultra Services Platform while Samsung was responsible for the virtualized RAN, base stations and home routers.
And on the 3GPP side, the remit is already broadening, with Qualcomm leading a study to run 5G NR in unlicensed spectrum. Qualcomm – which also drove the push of LTE into unlicensed bands with LTE-Unlicensed, LTE-LAA and MulteFire – said the initiative would look at many bands including 60 GHz, the millimeter wave spectrum currently used mainly by WiFi derivative WiGig.
Rasmus Hellberg, senior director of technical marketing at the chip giant, said in an interview: “3GPP included Listen-Before-Talk (LBT) in LTE-LAA and worked diligently to ensure fair sharing with WiFi in 5 GHz, which will serve as the foundation for the 5G NR in unlicensed study. Therefore we are confident that NR in unlicensed bands will also coexist fairly with Wi-Fi when they share the same unlicensed spectrum band.”
Hellberg wrote in a blog post: “The 3GPP has approved a study of 5G NR in unlicensed spectrum, both with an ‘anchor’ in licensed spectrum (the 5G LAA version) and standalone in unlicensed spectrum (the 5G MulteFire version). In other words, NR in unlicensed will become the 5G evolution path for both LAA and MulteFire. This study will eventually result in expanded functionalities for 5G in future 3GPP releases that are part of our 5G NR Spectrum Sharing vision, or NR-SS for short. We still have more work to do, but we are working meticulously to ensure that 5G NR will support all spectrum types, including new exciting spectrum sharing paradigms enabled by NR-SS, so that a wider ecosystem can enjoy the benefits of 5G.”
Nokia tests 28 GHz 5G with KDDI:
Nokia’s latest 5G trial is with KDDI of Japan, in the voguish 28 GHz band. The trial in Fujimino City used Nokia’s AirScale radio and achieved speeds of 1Gbps between KDDI’s research building and an apartment complex 100 meters away.
The trial focused on high speed broadband for built-up areas and Jae Won, head of Nokia Japan, said: “This trial is an important milestone both in our collaboration with KDDI and in our development of technology to meet the ever-growing demands of the large populations of megacities. This trial is a great example of how it can be applied to transform the entertainment experience as one of the many use cases 5G has to offer.”
Both companies signed a memorandum of understanding last year to research 5G technologies. KDDI rival Softbank has also been testing in 28 GHz, with Ericsson and Huawei.
Ericsson’s latest 5G boasts come from South Korea, where it has been building and optimizing a test network for KT, in preparation for the 2018 Winter Olympics in PyeongChang.
Ericsson and KT have also conducted tests along the Incheon Airport Railway—a new high speed rail line linking capital Seoul with Incheon International Airport. The tests included a handover between 5G base stations along the railway and a peak throughput of 4Gbps. The two companies have agreed to jointly develop beamforming/beamtracking technology and 5G-LTE interworking technology to optimize capacity and coverage.
KT is already in the process of conducting interoperability tests based on the PyeongChang 5G SIG specification in order to have pre-5G systems and services available by October this year. Nokia is also working on 5G with KT.