Now that more than 50 service providers on every continent have rolled out 5G, where are we with 5G? and where we are going? are timely questions. When will we get to 5G Standalone (5G SA)?
To date, all 5G networks have been launched with dual connectivity (DC) architecture. DC means that the User Equipment (UE) is connected to two base stations at the same time. In the 3GPP standards, four DC configurations are defined. (See Figure 1 for all 5G configurations)
The DC configuration chosen for the launch of today’s 5G networks is commonly known as Option 3x. Formally, it is known as E-UTRA-NR Dual Connectivity (EN-DC). The EN-DC configuration is defined as a 5G Non-standalone (5G NSA) architecture. An EN-DC configuration relies on the 4G Evolved Packet Core (EPC) and an anchor 4G eNB base station that acts as the master node. In addition, the 5G NR gNB base station is tethered to the 4G eNB and acts as the secondary node. The DC configuration is actually quite clever, combining the capacity of both base stations. By leveraging the existing EPC, it allowed service providers to get to market faster, launching in 2019 rather than in 2020. This is where we are today.
Before we move on to what comes next, we need to look at where things stand with coverage. Today, 4G LTE gNB base stations are in low-band spectrum (< 1 GHz) and mid-band spectrum (1 GHz – 6 GHz). A rule of thumb about RF coverage is that if you double the frequency, you will get only half the radius resulting in one-fourth of the coverage of the lower frequency. This assumes that all other elements are kept constant, such as the air interface and duplexing, RF output power, order of modulation, antenna technology, and the antenna height. Of course, each of these variables will not necessarily be the same as we move into 5G. In the 5G mid-band spectrums at 2.6 GHz and 3.5 GHz, massive MIMO has been deployed, some with beamforming. This has stretched 5G coverage beyond conventional 4G deployments. To get maximum 5G coverage in the mmWave bands at 28 GHz and 39 GHz, lower modulation is used. This is something to keep in mind as service providers move 5G into higher frequencies. The struggle is to match 4G coverage with 5G at higher frequencies, both outdoor coverage and indoor coverage.
As a result, many service providers will find they are limited in 5G coverage as compared to the 4G anchor in the 5G NSA EN-DC configuration. The ratio of 4G coverage to 5G coverage will vary based on the difference in frequencies. Some service providers will be able to launch in the same frequency band as their LTE coverage (2.6 GHz) and will not have the same coverage problem, although it will limit the data throughput. Service providers that launch with mmWave coverage will have spotty 5G coverage upon initial launch, although they will have the highest data throughput (Figure 2).
The mid-band 3.5 GHz band has been promoted as a desirable 5G band because it is closer to the 1800 MHz band. But the rule of thumb that twice the frequency results in one-fourth the coverage means that the 3.5 GHz band would need up to four times as many base stations to match 4G coverage. Although probably less due to massive MIMO and beamforming, still, a 5G 3.5 GHz base station cannot match the 1800 MHz coverage of a 4G base station. This is the situation that many service providers find themselves in today: if they want to approximate the coverage of 4G with 5G coverage in 2019, they have only one choice—to stay with 5G NSA EN-DC configurations, combined with cell splitting. (Figure 3).
Other choices will present themselves in 2020 with migration to the 5G Core. More configurations of dual connectivity presents itself in addition to 5G SA. Instead of cell splitting, the 5G NSA EN-DC configuration can be upgraded to 5G NSA NGEN-DC (Option 7x) or 5G NSA NE-DC (Option 4), with 5 SA (Option 2) added to round out the coverage. If these are mmWave sites, we are looking at extremely high capacity. (Figure 4)
The 5G Core plays a key part in enabling 5G Standalone, which allows service providers to round out their 5G coverage. The 5G Core is the main facilitator of multi-access edge computing (MEC). Service providers will be able to distribute the user plane function (UPF) to locations that meet users need for lower latency. Regional data centers will offer 10-20 ms latency and edge and deep edge data centers will offer latencies of 1 to 10 ms. Combined with network slicing, service providers will be able to custom tailor services for different vertical segments, as illustrated in Figure 5.
The launch of these features will take time. Starting in 2020, we will see the advent of 5G Standalone networks empowered with the 5G Core, as service providers expand their 5G footprints. This will eventually allow service providers to deliver on all of 5G’s promises: (1) enhanced mobile broadband (eMBB), requiring both extremely high data rates and low-latency communication in some areas with reliable broadband access over large coverage areas; (2) massive machine type communications (mMTC), supporting up to one million devices per square kilometer; (3) ultra-reliable low-latency communications (URLLC); and (4) fixed wireless access (FWA).
Dave Bolan is the senior analyst for Dell’Oro Group’s packet core research. He is author of the Wireless Packet Core five-year forecast and quarterly vendor share reports, and co-author of the upcoming Multi-access Edge Computing report.
