2021 was a crucial year for the acceleration of 5G, following the initial network launches in 2019 and 2020. There is even more news in store for us in 2022 as the 5G landscape continues to evolve. AT&T, and Verizon will launch their C-BAND spectrum this year, bought collectively for $68.8 billion in 2021. My colleague Will Townsend and I talked about the impacts of the C-BAND auction and its rollout delays in our G2 on 5G Podcast over the last year. While the implications of that auction are too vast to cover in this article thoroughly, it will significantly impact 5G’s future in the US. This will be a multi-part series, starting with the state of 5G networks, followed by 5G devices, and finally 5G use cases and services. So buckle up—it’s going to be a fun ride.
2022 will be an even busier year for operators than 2021
Let’s start with C-BAND. This spectrum range is critical to the US 5G rollout for two reasons. First, it represents the country's largest available swath of mid-band 5G spectrum. Second, the mid-band spectrum delivers a better mix of speed and coverage than mmWave and low-band. 2022 began with the conclusion of the FCC’s C-BAND auction, which raised $81 billion in total for the federal government. The FCC initiated this auction, which, at the time, prioritized the $81 billion over enabling innovation. There hasn’t been a spectrum auction in the world that has ever cost carriers anywhere near $81 billion; the next closest auction would be the AWS-3 auction which netted the US government $45 billion in 2015, and back then, that was a big deal. In addition to the cost of the spectrum, the operators have to pay for clearing costs which get the C-BAND total up to $95 billion. When the operators have to spend so much on spectrum, it ultimately gets passed onto the users, whether consumers or businesses. Because in addition to spending $80 billion, the operators will have to spend tens of billions rolling out that spectrum by upgrading existing cell sites and installing new ones. The FCC, a government agency appointed by Congress, regulates all radio waves, which is how it became the regulator in charge of auctioning spectrum for cellular operators to use for 5G.
In October 2021, only two months before AT&T and Verizon were initially set to launch their C-BAND spectrum networks; the FAA blew the whistle over C-BAND’s potential to interfere with plane ground radar. The FAA (Federal Aviation Administration) – part of the Department of Transportation, which regulates air travel and aircraft claims that ‘spurious emissions’ coming from 5G cell towers could interfere with the ground radar altimeters on aircraft. This is even though there have been no reported cases of commercial 5G networks or equipment causing any interference. Additionally, the C-BAND spectrum that the FCC is releasing is the furthest on the radio spectrum from the 4.2 GHz equipment that the FAA is concerned about related to 5G. The closer bands to 4.2 GHz, the B and C blocks of the C-BAND spectrum won’t be available until December 2023. Many in the industry, myself included, believed these concerns to be politically motivated. AT&T and Verizon paid top dollar for the lowest frequency and earliest available spectrum (100 MHz A-Block from 3.7 to 3.8 GHz), assuming that this would not be an issue and it hasn’t been in Europe, not to mention countless commercially available filters exist that render this argument a moot one. This spectrum features a 400 MHz buffer from the 4.2 GHz commercial airliners used for air-to-ground radar. These possible interference issues can easily be mitigated with numerous already commercially available filters. The rub is that much older aircraft lack these filters, as I see it. The FAA and its stakeholders, the aerospace industry, weren’t involved or included in the initial process of selling this spectrum. The organization and its stakeholders hope to get their piece of the $80+ billion to pay for these filter upgrades.
T-Mobile shelled out $9 billion for its share of the coveted C-BAND spectrum in the same auction as Verizon and AT&T. However, it bid on less spectrum and in fewer areas because the 2.5 GHz mid-band is already the bedrock of its 5G network, with almost 200 MHz of spectrum in some markets. Consequently, T-Mobile is far less affected by the FAA directives and other issues around C-BAND, given that its C-BAND spectrum likely won’t roll out until 2023. The company began rolling out its mid-band 2.5 GHz network, the ‘Ultra Capacity’ band, in April 2020, with spectrum gained in its acquisition of Sprint. This, along with the company’s low-frequency, broad-coverage 600 MHz band, its C-BAND spectrum, and tiny patches of mmWave (28 GHz), will comprise the bulk of T-Mobile’s 5G network for the foreseeable future.
What this means, essentially, is that T-Mobile has a head start on 5G deployments; its low-band 600 MHz network currently provides the best 5G coverage in the US, and its mid-band deployment has a two-year head start on the competition. This past November, T-Mobile’s mid-band ‘Ultra Capacity’ network purportedly achieved the coverage milestone of 200 million Americans.
In a personal anecdote, last month, I ran a speed test on T-Mobile’s UC 5G network while sitting at a stoplight in suburban San Diego that clocked in at 1 Gbps. I wasn’t expecting to see those kinds of speeds for quite some time, outside of, perhaps, New York City, T-Mobile’s most robust and most mature market. It struck me as highly notable that I was able to see 1 Gbps on a commercial mid-band 5G network in a suburban environment.
AT&T has quietly allowed Verizon and T-Mobile to duke it out over 5G claims, hoping the world will forget the company’s earlier efforts to rebrand 4G LTE Advanced as 5Ge. This was theoretically done to compete with T-Mobile’s 600 MHz 5G network, which was set to immediately overtake the speeds available on AT&T’s initial mmWave 5G deployment in 2019. That said, while AT&T’s 5G coverage is nowhere near T-Mobile’s today, it is still quite good. While the speeds are only slightly better than 4G, they do keep up with T-Mobile’s low-band offerings. Still, the release of the C-BAND spectrum will be AT&T’s first real opportunity to take advantage of 5G’s much faster speeds and lower latency. AT&T doesn’t have a standalone 5G network quite yet, but I suspect it will once mid-band reaches a certain level of coverage and speeds. Based on what we had seen from AT&T in 2021, I would expect the company’s deployment of 5G to land somewhere between Verizon and T-Mobile’s in terms of total coverage and speeds.
I think Verizon’s fixation on having mmWave 5G as the core of its network was a misguided one and is why it ultimately why the company had to spend $45 billion on C-Band 5G spectrum. There was no other way to compete with T-Mobile and AT&T moving forward. Verizon’s projection that its mid-band network will cover 200 million Americans by the end of 2023 is interesting. The company also says the initial 100 million will be covered by the end of the first quarter of 2022. Taken together, the initial launch looks pretty rapid, and the buildout, surprisingly slow and gradual.
In my experience, 200 million is the threshold for good 5G coverage and user experience. When all three carriers have coverage in that ballpark, I believe we will see 5G use cases take off and developers able to take advantage of the higher speeds and lower latencies truly. I believe that T-Mobile will continue to lead the competition for the coming years, with AT&T and Verizon starting to meaningfully close that gap in 2023 as long as the FAA and aviation industry don’t push back the launch of C-BAND even further. We need AT&T and Verizon to be competitive with T-Mobile for the US to drive innovation and become a genuinely competitive 5G market that breeds new use cases and industries.
In addition to C-BAND (Auction 107), we also just concluded the 100 MHz block of 3.45-3.55 GHz spectrum, which is known as Auction 110. This auction came in at $22 billion, with AT&T spending $9 billion, Dish spending $7 billion, and T-Mobile spent $3 billion. This band (n78) is essential because it can globally harmonize with other operators worldwide and could enable a whole host of devices with global 5G support using a single band. This auction also gives some relief to those who might not have wanted to participate in the C-BAND auction that cost upwards of $81 billion. $22 billion is still the third biggest auction in the FCC’s history. I believe the third biggest in the world, so its still nothing to scoff at, but I believe that AT&T and T-Mobile are both using it to shore up their mid-band holdings while Dish will likely be using it as a core for its own 5G network’s capacity since it currently mostly holds low-band spectrum today. US Cellular also spent $580 million but paid much less per license than the other four bidders who spent more because most of their licenses are likely in rural areas where US Cellular traditionally plays. AT&T paid $5.5 million per license, Dish paid $5.9 million, and T-Mobile paid $14 million, all in stark contrast to US Cellular’s $1.5M and Columbia Capital’s $77 million. The disparity between these license prices is predominantly based on the desirability of the licenses, meaning that spectrum in more populated areas is likely to sell for more than spectrum in less populated areas. T-Mobile is only buying spectrum in places that it doesn’t think its 2.5 GHz or C-BAND holdings will be enough, while Columbia Capital is speculating that the spectrum it has acquired will appreciate or it will be able to lease spectrum to an operator like it did with 600 MHz with T-Mobile.
Outside of the US, we saw different European operators build out their 5G networks. Operators in the Middle East propelled countries like the UAE and Saudi Arabia towards the top of average mobile speeds on SpeedTest and OpenSignal. China’s biggest operators, China Telecom, China Unicom, and China Mobile, have aggressively added new 5G customers monthly. The Chinese MIIT claims the country now has well over 700 million 5G subscribers and over 1.3 million 5G base stations. China Mobile claims 374 million 5G subscribers, up from 150 million in November of 2020. China Unicom claims 150 million 5G subscribers, up from 78 million at the beginning of 2021. China Telecom now claims 178 million 5G subscribers, purportedly adding over 92 million subscribers in 2021. These numbers appear to significantly surpass various industry forecasts for 2021, including Qualcomm’s projected 450 to 550 million 5G handsets sold by the end of the year.
India is arguably the biggest laggard when it comes to 5G, which isn’t much of a surprise when you consider how long it took the country to get 4G. India didn’t launch 4G (via Bharti Airtel) until 2012, a whole three years after Verizon’s initial launch of 4G LTE in 2009. It’s almost fitting that India will again likely be three years behind Verizon (which launched its first 5G network in 2019) when it launches its 5G network in 2022. The auction for spectrum is expected to happen sometime in April to May, with the launch of 5G commercial services expected on August 15th, India’s Independence Day.
However tardy, the Indian market’s adoption of 5G will likely be a significant accelerant of 5G’s uptake in 2022, given that India has the second largest mobile market in the world (roughly 1.5 billion mobile users). I expect that in 2022 and 2023 we will see a second phase of 5G growth, with improved coverage and speeds as the more developed 5G markets convert the last of their holdout users and new markets like India explode. I also believe that India’s entry into 5G will herald a huge new swath of mid-tier and entry-level 5G devices, like we saw with 4G. I do believe that upon launch India’s 5G uptake will likely be faster than it was with 4G, especially when you consider India’s population density in urban areas where most users reside. We’ve seen this trend play out similarly in other countries’ rollouts.
It will be interesting to see how much uptake India’s home-grown 5Gi standard experiences, or if it dies on the vine. If the Indian government wants to improve 5G coverage outside of urban areas, it will need to either help subsidize the cost of 5Gi equipment or incentivize more rural deployments. While the Indian government planned to mandate 5Gi as its own standard, it remains to be seen how much of 5Gi will be adopted by infrastructure vendors. The 3GPP has already agreed to integrate 5Gi into the Release 17 standard, expected sometime in March or June of this year. This will enable hardware to implement 5Gi capabilities as soon as 2023—appropriate since India’s 5G launch won’t even begin until the second half of 2022. That said, there is still no clear indication of the added cost of supporting 5Gi and how much it will impact coverage.
In Europe, according to the 5G observatory, there are 25 countries with commercial 5G services. Europe also has twelve 5G corridors that allow for seamless connectivity across countries, spanning road, train, and maritime routes. 2021 saw several European 5G spectrum auctions, including Spain (700 MHz), Croatia (700 MHz, 3.6 GHz, and 26 GHz), and Portugal (700, 900, 1800 MHz, 2.1, 2.6, and 3.6 GHz). As of December 2021, 53% of Germany has 5G coverage, thanks to the 3.6 GHz spectrum and DSS (Dynamic Spectrum Sharing), a technology that enables the sharing of 4G and 5G on the same band. Verizon has also utilized DSS to improve its coverage in the US, as it competes with T-Mobile and AT&T’s 5G coverage ahead of the C-BAND launch.
It’s notable that Europe has not had the same concerns around spectrum interfering with aviation and hasn’t faced the same rollout delays. This fact throws the FAA’s story even further into question. The same holds in Japan, the first country to deploy 3.8 to 4.2 GHz for 5G back in 2019, and the UK, which followed shortly after that. To date, there have been no reports in those countries of issues around airports that have 5G. Granted, Europe is primarily using 3.6 to 3.8 GHz, but that is close to the US’s C-BAND spectrum (3.7 to 3.98 GHz) and still at least 200 MHz away from the 4.2 GHz frequency used for radar. There is also a 20 MHz guard band between C-BAND and the re-packed incumbents that sit on 4.0-4.2 GHz.
Japan, South Korea, and Australia have their own 5G deployment strategies. Lagging somewhat behind South Korea and Australia, Japan has four operators offering 5G: NTT DoCoMo, Rakuten, and Softbank. 5G availability in Japan is still not great in terms of overall coverage. Rakuten offers the best average download speeds (224.3 Mbps, but has the least reach of any operator and isn’t available in many areas. What makes coverage so difficult in Japan is the fact that the country does not have any low-band 5G spectrum. Japan’s currently available bands are 3.6 GHz to 4.1 GHz, 4.5GHz, and 28 GHz, all of which have a more challenging time propagating than lower bands. Japan’s coverage problem could potentially be solved by introducing more open-source, low-cost commodity equipment. It appears that Rakuten is taking this tack as part of its Symphony business, which it plans to offer around the world. Leveraging Open RAN, Symphony's goal is to accelerate network deployments, make them more cost-effective, and promote competition. Ideally, this should enable faster and cheaper coverage across Japan and other regions. Future plans aside, Japan’s coverage is still fairly abysmal. Hopefully, 2022 offers a change for Japanese 5G users.
On the other hand, South Korea’s 5G market has maintained an aggressive pace of growth—27% of its total cellular subscribers are now on 5G services. Korea is primarily governed by three major operators: SK Telecom, KT, and LG U+. All three offer an average of 5G download speeds ranging anywhere from KT’s 367 Mbps, LG U+’s 414 Mbps up to SK Telecom’s 467.4 Mbps. 5G availability in South Korea is also considerably better than in Japan, with LG U+’s coverage at 26.3%, SK Telecom at 30.8%, and KT at 28.6%. In OpenSignal’s methodology, availability does not equate to coverage but rather what percentage of the time that users have any kind of a connection. Meaning that a 5G device was on the 5G network in that percentage of time on the operators network, whether it is 3G, 4G or 5G. The majority of South Korea’s 5G coverage and speeds come from the roughly 100 MHz of spectrum each carrier has in the C-BAND range, though the country does have some mmWave spectrum available (26 GHz to 29 GHz).
Altogether, Korea still has the highest 5G availability of any country in the world, according to OpenSignal’s November 2021 report. This number is higher than much smaller countries, including Kuwait, Hong Kong, Qatar, and Taiwan. That said, I believe that South Korea could benefit greatly from more 5G mmWave deployments, given the population density of its bigger cities and urban areas. In a recent trial, Samsung demonstrated the potential benefits of using mmWave 5G. When implemented as a backhaul for subway Wi-Fi, mmWave increased download speeds by 25 times. Wi-Fi download speeds peaked as high as 1.8 Gbps.
In Australia, you have three main 5G operators: Telstra, Optus and TPG. Australia has the difficult task of addressing a country in which 85% of the population lives within 50 km of the coast. One thing this means is that when Telstra claims it covers 75% of Australians with 5G service, it really doesn’t have to cover much ground to do that. Neither do any other operators in Australia. The problem is that Australia itself is enormous and there’s a large disparity between the distribution of cell sites in the major cities and vast, sparsely populated rest of the country. While Telstra is starting to explore the use of mmWave 5G in densely populated central business districts in bigger cities, it still heavily relies on low-band 850-900 MHz and mid-band 3.5-3.7 GHz for the bulk of its networks. There was a recent auction in December of 2021 in which both Optus and Telstra bid on low-band spectrum to help with coverage. TPG did not participate in this auction as it has already deployed its 700 MHz holdings as a standalone 5G network. If T-Mobile’s 600 MHz standalone 5G network is any sign of what’s possible with a standalone low-band 5G network.
In addition to the state of networks and operators, we will be talking about devices and future applications and use cases. That’s where we will be able to dig into even more about the future of 5G and where the standards bodies and companies driving those standards bodies expect the technology standard to go. Devices are only now starting to catch up to where the standards are today, while operators are clearly struggling to keep up with the pace of innovation, some more than others. In part two, we will talk about devices and chipsets driving those devices and what uptake has been for 5G devices as well as where we expect the future of 5G devices to go.