Spectrum has been the lifeblood of the wireless industry since its inception. Without it, mobile connectivity simply doesn’t exist. That remains true as ever in a 5G world, but 5G also dramatically changes the breadth and depth of spectrum that is required to power lower latency, higher bandwidth, and more connections per cell site.

The spectrum held by most operators today is not enough to deliver all of these key features of 5G, and early 5G customers have experienced these limitations first hand. Operators will need hundreds of megahertz of clean spectrum to provide 5G networks sporting the oft-promised multi-gigabit speeds.

Network operators will, in most cases, need a healthy mix of low-, mid-, and high-band spectrum to fulfill that promise. None of America's mobile network operators have the right ingredients to mix a genuine 5G cocktail today. 

That reality is made all the more complex because of the unique characteristics of these different bands of spectrum at play in 5G networks. For every positive associated with low-, mid-, and high-band spectrum, there are just as many negatives.

While some rapidly advancing technologies like dynamic spectrum sharing, network slicing, orchestration, and virtualization can minimize the effects of these challenges, the laws of physics still apply. Technology has its limits, and operators have always adapted to and worked within those guidelines.

There is a widely agreed categorization for these three bands although there are some delineations and nuances that operators prefer to highlight, depending on their spectrum position. 

Low-band spectrum includes everything below 1 GHz; mid-band spectrum includes all frequencies between 1 GHz and 6 GHz; and high-band spectrum generally includes all frequencies above 6 GHz, though with a current focus between 24 GHz and 40 GHz. High-band spectrum also has the special privilege of being referred to as millimeter-wave (mmWave) spectrum.

It is important to take note of the different spectrum bands at play for 5G because U.S. operators have embraced very different strategies for 5G deployments based on the bands for which they have licensed spectrum, according to Dan Hays, principal at PwC’s Strategy& division. 

“It’s very important to understand the differences because it’s probably going to be quite apparent at least for the next several years where there are some stark differences in the network characteristics of different operators purely based on the spectrum that they own,” he said. 

“Are the buckets as neat as those three categories make them sound? No, certainly not,” Hays said. “There’s some very fuzzy lines between them, but it is a useful categorization. … The trials and tribulations of each of the bands are very real.”

Get Low, Stay Low

Low-band spectrum is also sometimes referred to as beachfront property because of its ability to propagate far distances and penetrate buildings and foliage. But those properties also make it very desirable for a lot of needs, which means there is not a lot of it available specifically for telecommunication networks and thus not a lot of capacity.

Mobile network operators have been using low-band spectrum for decades, going back to the earliest wireless networks. While that provides the comforts of familiarity, low-band spectrum is also heavily congested. T-Mobile US’ 5G network is a prime example. 

While the operator’s 600 MHz spectrum holdings enabled it to deploy the first nationwide 5G network in the country in late 2019, its licenses in that band are also relatively narrow. As a result, T-Mobile has widespread 5G coverage reaching a potential of 200 million people, but speeds are nowhere near the level expected of a 5G network.

“T-Mobile sort of revolutionized the segment to some extent because they have the coverage of 200 million [potential customers],” said Jeff Moore, founder of Wave7 Research. Low-band 5G “speeds are not as revolutionary as one might have expected but the coverage is quite good.”

The near-term outlook for additional low-band spectrum remains poor, according to Hays. He expects the industry to coalesce around another effort to reallocate low-band spectrum currently used by TV broadcasters for more cellular capacity in the next couple of years, but it’s unclear when and how much spectrum that might free up for mobile operators.

“The quandary for mobile network operators is how to refarm the existing low-band spectrum that they have for use with 5G,” he said. For most operators that means shutting down 3G networks that are operating in low-band spectrum, but that too will take many years to materialize. 

Operators Covet the Spectrum Middle Ground

Mid-band spectrum is “very much the focus these days” because it provides a “mixture of coverage and capacity,” Hays said. However, in most cases it still “requires significantly more cell sites than are in place today in order to have equivalent coverage,” he added. 

Sprint’s massive spectrum holdings in the 2.5 GHz band factored heavily into T-Mobile’s decision to acquire the struggling operator. So much so that T-Mobile has been battling to win approval to buy Sprint for two years. The culmination of that quest is now imminent.

Sprint put some of its 2.5 GHz spectrum to use last year when it activated 5G service in 11 markets covering approximately 16 million people. T-Mobile plans to put that spectrum into use quickly after the deal closes. 

AT&T’s 5G network riding on low- and mid-band spectrum currently covers 80 million people across roughly 80 cities and the operator says it will reach nationwide availability before July. The operator is also planning to sunset its 3G network running on mid-band spectrum, but those efforts are ongoing. Verizon’s position and outlook for mid-band spectrum is less clear.

“Mid-band spectrum has perhaps the greatest potential for 5G given the balance between coverage and capacity,” Hays said. “The good news for network operators is that there is a large amount of mid-band spectrum that is either now available or slated to become available, but it still may take several years for all of them to really make its way into networks.”

The Citizens Broadband Radio Service (CBRS) band, recently approved by the Federal Communications Commission (FCC) as a shared spectrum platform riding in the 3.55-3.7 GHz band, falls into this category. CBRS is largely expected to be used for in-building coverage, public spaces, and industrial IoT applications.

5G Operators Are High on High-Band

High-band spectrum, while not necessarily new, is a recent phenomenon and area of heightened interest for mobile network operators and equipment vendors. It can deliver speeds above 1 Gb/s, but its propagation characteristics are woefully inadequate for the vast majority of users.

“There are just lots of challenges with building out strong coverage using millimeter-wave” spectrum, Hays said. While mmWave spectrum is “probably a terrific solution” for high-density areas, it is “unlikely to be the long-term solution for suburban and rural areas that are less dense, and have less availability of fiber networks needed to transport” the traffic.

Verizon, which has thus far focused exclusively on mmWave spectrum for its 5G network, has encountered many of these challenges. Researchers and field tests have indicated that mmWave signals can drop within 1,000 meters of a small cell, rendering the reach of those sites even more narrow than initially expected.

Verizon maintains that it can overcome those challenges with beam-forming technology and a higher concentration of small cells, but that also requires higher costs and the permits required to deploy an abundance of small cells.

Because mmWave is playing a starring role in 5G it will require a 10- to 100-fold increase in the number of cell sites spread across the country. The top of that range equates to a potentially astounding figure of 30 million cell sites around the country, but the number could reach even higher depending on the performance of mmWave as deployments spread. 

Indeed, mmWave spectrum bands “have been horrible for propagation and actual coverage,” Moore said, adding that the percentage of population covered by Verizon’s mmWave 5G network to date “is extremely tiny.” AT&T and T-Mobile also hold licenses for mmWave spectrum and have deployed 5G in some markets using that spectrum to mixed reviews.

As it stands today, the mmWave network performance “from Verizon is very much ineffective because the 5G that they have is more theoretical than real to 99% of the population and 99% of the cases,” Moore said. Looking ahead, however, he expects dynamic spectrum sharing to help mitigate some of those challenges because it will help operators blend coverage across multiple spectrum bands, and provide better coverage as a result. 

Spectrum Complexities Muddy 5G Promise

While spectrum has always been critical for mobile networks, it’s “arguably even more important in a 5G world where most existing spectrum is already in use typically for 4G,” Hays said. “There’s precious little room to maneuver in building new networks to shift usage from 4G to 5G.”

Considering the various and complex spectrum strategies of U.S. operators for 5G, the industry is “doing the end users and enterprise users a disservice by labeling anything as 5G,” he said. “The differences in networks are so stark, in many cases, that we would be better off having a different lingo to describe the capabilities of the network instead of painting them all with the same broad brush of 5G when the user experiences may well be vastly different.”

Hays argued that the industry would benefit from some “meaningful terminology that really identifies the capabilities of various 5G networks, and today I’m not sure anyone has really cracked that code.” As such, Hays isn’t convinced that U.S. operators have the spectrum required to deliver on the promise of 5G. 

“The road to 5G for U.S. mobile network operators is actually quite unclear, and there are two main constraints. One is spectrum availability where many of them just don’t have the fallow spectrum to deploy 5G everywhere, and then the second is the availability of capital,” he said. 

Unless 5G can deliver an uplift in revenue, and that’s a big if, “we believe that it will actually wind up being a fairly slow rollout that will concentrate initially on only the most dense urban areas and really move slowly over a period of probably at least five more years from now before we get a reasonably widespread deployment of 5G,” Hays predicted.

These challenges arrive every decade when operators introduce new network technologies, and they usually get sorted out over time. But for now, the prospects for 5G and the spectrum required to deliver all the promises of 5G is messy.