Verizon C-Band 5G: We Tested it. It Works!
Sometimes, simple demos are the best. The short version of this report is that Verizon flew a handful of press, influencers, and analysts to downtown Los Angeles last week, gave us phones with access to a test C-Band network set up around the L.A. Live venue, and let us see how fast it was. TL;DR: It worked! An unloaded C-Band network with seven people using 60 MHz of spectrum is really fast. For details, what this means, and why it matters, read on.
Context: U.S. 5G Frequency Band Deployment
5G is being deployed in three different areas of the spectrum map: low-band, mid-band, and mmWave. In most of the world, 5G launches were clustered in mid-band frequencies, which travel reasonably far and through structures for some indoor coverage. Aside from Sprint, which accumulated a hoard of 2.5 GHz mid-band spectrum it intended to use for 4G, those frequencies were in use by satellite operators in the U.S., so the FCC chose to auction off huge swaths of mmWave spectrum first. AT&T, T-Mobile, and Verizon all bought mmWave holdings, but Verizon was the most aggressive in deploying it. mmWave offers truly dizzying speeds and huge capacity, but at the cost of propagation: it doesn’t go far and can’t pass through walls, people, tree branches, or sometimes even leaves. That makes mmWave perfect for open heavily trafficked areas like Times Square and stadiums, but uneconomical and impractical to use everywhere. U.S. carriers all rolled out limited 5G service on their existing low-band networks by carving out space and shutting off the 2G networks, but there just isn’t much room on those spectrum slivers, and 4G networks are often faster.
Why Carriers Need C-Band
In February 2020, the FCC hatched a plan to clear 280 MHz of C-Band spectrum (3.7 – 3.98 GHz) in two stages, and later that year auctioned it off. In the meantime, T-Mobile finally finished buying Sprint for its 2.5 GHz mid-band spectrum in April 2020, so when the C-Band spectrum was auctioned off later that year, it only needed to “top off” its 2.5 GHz network with $9.3 billion of C-Band licenses. By earlier this month, T-Mobile had finished rolling out 5G service on Sprint’s 2.5 GHz mid-band spectrum that reaches 200 million people. In our testing across various New Jersey suburbs, we have seen download speeds range from 235 Mbps to 550 Mbps – much higher than most wired broadband connections in the U.S. This presents a serious competitive challenge to rivals. Verizon can point to its mmWave 5G in some stadiums and city blocks, and in my testing at MetLife Stadium during the only good Jets game this year in October, I saw 1.43 Gbps peak downloads in the stands and 891 Mbps downloads on the field after the game. Still, depending on where they work and live, most consumers may never encounter mmWave networks at all.
For that reason, AT&T and Verizon spent big on the 2020 C-Band auctions to build their own fast-and-accessible mid-band networks. AT&T spent $23.4 billion before clearing costs, but the biggest winner was Verizon at $45.4 billion. Verizon won at least 140 MHz of C-Band across the U.S., with 60 MHz of that clearing by January 5, 2022. Verizon also won 200 MHz in 159 mostly rural areas covering 40 million people; in addition to smartphone service, Verizon intends to use this spectrum to expand fixed wireless offerings and compete with local broadband providers. Clearing costs and network buildout will add tens of billions of dollars to the tab. Verizon has been aggressively installing new equipment and intends to turn on its C-Band network starting on January 5 and provide service to at least 100 million people by the end of Q1 2022. But that leaves some simple questions: does it work? What will that network look like for users? Also some more complex questions: will it be competitive with T-Mobile’s mid-band 5G network at launch? How about down the road?
Testing Verizon’s C-Band Network
To answer that question, Verizon lit up a test network in downtown Los Angeles using the 60MHz of frequency that will officially be theirs next month, and flew a few of us out to test it with real-world devices. All the test network antennas were outdoors; any indoor coverage originated externally. Verizon provided a pair of Samsung Galaxy S21’s tuned to either see just the C-Band network and low-band, or the new C-Band network along with low-band and mmWave. I brought along my own Google Pixel 6 Pro with a Verizon eSim, an iPhone 13 Pro on T-Mobile, and a Microsoft Surface Duo 2 on AT&T. We were given the location information and phones just moments before we went out to test the network, so I did not have a rigorously planned route or test protocols, and I did not test all the phones in every spot. That said, I was still able to get some useful data and impressions.
Downloads Are Fast
In general, when you are on an unloaded mid-band network, it’s going to be fast: I saw download speeds averaging 725 Mbps near the transmitter at the Christmas tree, speeds of 600 Mbps inside a glass-walled fast food store nearby, and downloads around 200 Mbps at a location down the block and around the corner from the transmitter. This was still not as fast as Verizon’s mmWave network, which hit 1.2 Gbps down in the first location, 1.4 Gbps inside the restaurant when facing the window, and 1.7 Gbps in the location around the corner (there was clearly a separate antenna nearby that location).
The impressive mmWave indoor result on Verizon depended on a line of sight to the transmitter; when I faced away from the window or someone walked in front of me, service dropped all the way down to 119 Mbps on Verizon’s low-band network. I also found that the multi-band phone test phone prioritized C-Band over mmWave when it was near the transmitter. That’s not supposed to happen, and Verizon chalked it up to a configuration error (one of the Verizon engineers quipped, “thanks for finding that -- that’s why we set up these test networks.”). In the future, Verizon will have carrier aggregation across both its C-Band and mmWave spectrum. The bottom line is that Verizon’s upcoming C-Band network should be able to provide a significant speed boost over its current low-band 4G and 5G networks with reasonable signal penetration inside and around buildings.
Uploads Are Not Especially Fast – On Purpose
Upload speeds were considerably more limited, ranging between 45 Mbps up when away from the tower and 75 Mbps at the ice rink. When I could see the mmWave network, I also got 75 Mbps uploads. Verizon says that C-Band uploads are happening on the C-Band network itself (rather than using low-band or 4G), and they are artificially capped at those speeds because that’s what the market demands – super-fast downloads. We’ll see if this changes over time with people livestreaming and uploading larger photo and video content; Verizon says it can adjust C-Band upload speeds as needed.
The Competition
Comparisons to AT&T and T-Mobile at L.A. Live were interesting. AT&T did not seem to have mmWave deployed in the plaza, and its low-band 5G speeds were disappointing: 48 Mbps down and 28 Mbps up. T-Mobile was more instructive. At the ice rink, I got 303 Mbps down and 49 Mbps up – not as fast as Verizon’s unloaded, right-next-to-the-transmitter performance, but certainly solid mid-band numbers. Around the corner where Verizon’s C-Band dropped to 200 Mbps, T-Mobile hit an astounding 1.1 Gbps down and 47 Mbps up – I didn’t see one, but I can only surmise that T-Mobile had a 2.5 GHz antenna right across the street. Of course, if you had a Verizon 5G phone at that location, it wouldn’t necessarily be using C-Band anyway, it should latch on to the mmWave network for 1.7 Gbps.
Verizon’s C-Band network will start out with 60 MHz of spectrum and grow to 140 MHz or even 160 MHz in some markets in 2023. AT&T starts out with 40 MHz of C-Band, growing to 80 MHz in some large metro areas in 2023, which should place it in a similar position to Verizon, only with a lot less C-Band spectrum to play with initially and over time. T-Mobile’s mid-band network already has 100 MHz of spectrum in most markets and runs at a slightly lower 2.5 GHz frequency, which theoretically means that average speeds should be faster for the next couple of years, and achieving coverage should be easier. However, Verizon’s engineers are adamant that this is not a real-world problem. One representative said that coverage for Verizon’s C-Band (3.7 – 3.76 GHz) is the same as for 2.5 GHz, going to say, “I promise you it will be as good as T-Mobile despite any differences in 2.5 GHz vs. 3.7 GHz propagation.” Those are strong words for an engineer. Verizon also says that it plans to continue adding mmWave coverage in areas where combining C-Band and mmWave makes sense.
C-Band: Going Live January 5
While this was just a test network, Verizon will be starting its C-Band network rollout in earnest on January 5. Verizon will have a variety of devices that will work on the new network with software updates, with the first being Apple’s iPhone 12 and 13 families, Samsung’s Galaxy S21 family, and the Galaxy Z Flip3 and Fold3. Verizon was not willing to discuss the ongoing negotiations/blackmail with the FAA over airplane altimeter updates, but the CTIA insisted just this afternoon that the January 5 date will hold.
Who Cares?
Obviously, unloaded test network performance is not indicative of expected speeds once the network is deployed. I would not expect to see 750 Mbps downloads on the C-Band network at L.A. Live in a crowd after a Lakers game in 2022. Still, it’s nice to see what the upper bound expectations can be in the most favorable real-world conditions.
The bigger question is why anyone needs 300 Mbps speeds in the first place. A reasonable HD streaming video experience only requires a consistent 25 Mbps connection, so all this speed has limited utility today. The biggest initial benefit is simply capacity. 5G uses spectrum efficiently, and adding 60 MHz today, going to 140 MHz in 2023 can serve a lot of simultaneous connections. Verizon chose its test/demo location carefully: an outdoor plaza connecting a major concert theater (Microsoft Theater) and a sports stadium (the stadium formerly known as the Staples Center) that can fill up with thousands of people before and after events. This is exactly where you want as much network speed and capacity as possible.
Other benefits of extreme speed are quick downloads and streaming content. Verizon had us demo a live 3D feed from a 360 degree camera inside the stadium, and it access over C-Band was instantaneous and stable as I walked around the plaza. More utilitarian use cases are less exciting from a marketing perspective, but the ability to download and apply a security update quickly on the go, rather than waiting until you have a WiFi connection, can be meaningful, too.
Long term, there has never been an increase in computing power, storage, or network speeds that has not led directly to new application types and use cases. I’m confident predicting that at some point in the near future, a network that provides 300 Mbps downloads will be a requirement for something, whatever that something may be.
To discuss the full implications of this report on your business, product, or investment strategies, contact Techsponential at avi@techsponential.com or +1 (201) 677-8284.