The Six Paths to 5G Emerging in the USA

In a previous Foresight, “5G will Radically Transform RAN Design and Topology as Multiple Spectrum Bands and Use Cases Converge”, ABI Research discussed how spectrum and use case will fundamentally alter Radio Access Network (RAN) implementation in the 5G era. In this Foresight, we dive into the details of how the Mobile Network Operators (MNOs), Multiple System Operators (MSOs), and tower companies in the United States are implementing 5G. More background can also be found in our upcoming report 5G in the U.S.

Several mobile service provider executives have voiced their concerns that 5G revenue will grow at a much slower pace than that of 4G. This means that the new generation will be deployed in a phased manner and not nationwide—at least for the first years of its commercial life. Widespread adoption of 5G will be gated by user equipment and handset availability. At some point, the entire value chain will be conflicted. Will device manufacturers choose to add 5G capability in their smartphones when there is not wide coverage in the industry? Will the arguably limited consumer demand in the market justify chipset, device, and infrastructure advancements to make the technology commercially ready? An important part of this answer is the choice of 5G RAN implementation, and ABI Research outlines six alternatives planned by the U.S service provider community below.

Service providers must carefully capitalize on 5G over the next 10 years. There are many elements to 5G, and to expect them all to commercialize simultaneously is unrealistic. ABI Research expects 5G to start rolling out in 2019, with early deployments in carefully selected regions and cities in low- and mid-band spectrum, and to move to wide adoption from 2020 onward. mmWave spectrum will deploy in selected dense urban areas and regions in 2020 as the industry comes to terms with the challenges of propagation in these bands.

In the United States, the 5G plans announced by the proposed merger of T-Mobile and Sprint illustrate how the new T-Mobile will deploy 5G in complementary bands. Low-band will be the first to be deployed, using T-Mobile’s 600 MHz spectrum for nationwide coverage and the Internet of Things (IoT). Then, Sprint’s 2.5 GHz spectrum will be used for metropolitan areas for both 5G and Long-Term Evolution (LTE), but it will not be deployed nationwide as the capital expenditure requirements to do so would be prohibitive. The merged company will also deploy T-Mobile’s mmWave in select dense urban areas to augment speed and capability. The companies claim that this approach offers the best starting point for 5G and addresses all of 5G’s aspects by using spectrum across three bands.

This contrasts with Verizon’s plans that have deployed commercial pre-standard 5G (although not in the mobile market) in 28 GHz in five markets as Fixed Wireless Access (FWA) that offers featured content from Google/YouTube and Apple TV. The company expects to transition to mobile 5G New Radio (NR) in early 2019 by using network slicing, according to its Intelligent Edge Network Framework.

On the other hand, AT&T has planned commercial operations at 28 GHz and 39 GHz in 12 markets in 2018, followed by eight more in early 2019. The commercial service will feature content from DirectTV. This will be a true standards-compliant 5G; however, the terminal will be a “puck” hotspot that is manufactured by NETGEAR (not yet commercially available), which will provide the 5G connectivity and transcode the 5G signal to Wi-Fi. This will ease the transition to 5G handsets as the units become available. AT&T is building its network using Wireless Communications Service (WCS) (2.3 GHz), Advanced Wireless Service–3 (AWS-3) (1694–2180 MHz), and FirstNet (700 MHz); it is also ensuring that its towers are 5G ready at the same time. This is the company’s “one touch, one tower” approach—2,500 sites have been completed so far. Consequently, AT&T claims that the switch to 5G is a software upgrade.

The MSOs are entering wireless with a “quad-play” offering by leveraging the Mobile Virtual Network Operator (MVNO) contracts they have with the MNOs. The MSO commercial offerings, so far, are limited to Charter Communications (with its Spectrum Mobile 4G service) and Comcast (with its Xfinity Mobile 4G service). Altice USA is considering getting into mobile by bartering access to its network with Sprint in return for an MVNO agreement. Although none of these companies has deployed 5G, their deep-fiber penetration and hybrid fiber-coax networks will offer unrivaled capabilities for densification, backhaul, fronthaul, and compete head-to-head with the MNOs. ABI Research discussed this in our August Executive Foresight, “5G Emerging as Competitive Arena for Mobile Broadband, Cable, and Fixed Wireless.”

For its part, Dish Network is currently building phase one of its network, based on 4G Narrowband Internet of Things (NB-IoT) in 600 MHz and due for completion in March 2020. The second phase will be a 5G network based on the upcoming 3GPP Release 16 standard. That standard, which is called standalone 5G, or 5G SA, is intended to allow operators to launch a 5G network without requiring an underlying 4G network. According to Dish Network, this is likely happening in 2020, 2021, and 2022. Dish plans to act as a neutral host by leasing capacity on its standalone 5G network.

Tower companies—in addition to participating in 5G builds of the MNOs—can also leverage their assets to support edge computing infrastructure. Given the availability of power and backhaul at tower sites, compute, storage, and networking can be easily offered to reduce latency—a key component of the Ultra-Reliable Low Latency Communications (URLLC) and enhanced Mobile Broadband (eMBB) use cases in 5G. In the United States, Crown Castle is a bellwether towerco operating 40,000 towers, 60,000 small cells, and 65,000 route miles of fiber. The company is collaborating with VaporIO to augment the role of the tower as the industry disaggregates datacenter functionality to the edge of the network for ultra-low latency, virtualization, and hyperlocal content caching and distribution. The company’s Kinetic Edge platform plans dense compute per square foot at the tower sites or in wireless aggregation hubs as it anticipates the low latency demands of 5G’s URLLC. Also, SBA Communications, which owns and operates more than 28,000 towers in 13 countries and manages 9,000 towers on behalf of its customers, also plans to test the edge computing waters by working with Packet, a bare metal infrastructure provider, to deploy a datacenter at an SBA site in Foxborough (near Boston, Massachusetts).

We may be a witness to the cell tower becoming the new datacenter as we enter the 5G era.

In the United States, service providers will develop new RAN topologies as they monetize the URLLC, massive Machine Type Communications (mMTC), and eMBB use cases. The first steps have already been taken with Verizon’s FWA and AT&T’s hybrid 5G/Wi-Fi hotspot topologies. ABI Research counts at least six differing paths to 5G in the United States:

1. Verizon’s FWA and its transition to 5G NR in high-band
2. T-Mobile/Sprint’s triple band nationwide plan
3. AT&Ts nomadic 5G/Wi-Fi “hybrid hot spot” and mid-band plans
4. Potential MSO/MVNO networks from Charter Communications, Comcast, and Altice USA
5. Dish Networks 5G SA neutral host plan
6. Crown Castle and SBA Communications use of their tower sites as datacenters for URLLC and eMBB with perhaps other towercos to follow

Although there remain some fundamental challenges in 5G implementation, these differing approaches from the service provider community show that competition is fierce, as each company plans monetization of 5G in its own unique way.

Those companies that successfully meet these challenges will be the winners.