The Reality of Millimeter Wavelength for 5G

by Pasternack 

Amid the attention to the coming fifth generation of “cellular” technology, the use of spectrum at 28 GHz and higher is generating some of the attention. Most media coverage recites the well-known challenges associated operating at millimeter wavelengths and how it will require a huge number of small-cell case stations, massive MIMO, phased-array antennas, and the acts of high-tech magic to implement, especially within the confines of a smartphone. And it’s all true, which begs the question why any sane industry would perform this jump an order of magnitude or more in frequency.

There are actually very good reasons to tackle the millimeter-wave spectrum, which bears some explaining. When reading the following discussion, however, keep in mind that despite its inherent drawbacks, millimeter-wave spectrum offers tens of GHz of unoccupied spectrum, a virtual goldmine when compared with current lower-frequency allocations.

The Low-Hanging Fruit Has Been Picked

The spectrum most desirable for mobile operation is from about 100 MHz to 3 GHz. A glance at the ITU’s spectrum chart shows that this region is so densely packed with services that a magnifier is a handy tool for viewing each allocation. When a resource becomes scarce, imagination is required to use what’s left, and it’s far more difficult than simply picking a place to operate and deploying equipment there.

It takes spectrum sharing, refarming of current allocations, and other approaches. No one likes to share spectrum because of the potential problems from interference when two (or more) different services (like radar and wireless) are using the same frequencies. The way to keep this from occurring requires an enormously complex solution that has yet to prove that it works. And the result in every case is a few hundreds of megahertz or less of “new” spectrum.

Spectrum refarming is disruptive, costly, and time-consuming, but it has produced beneficial results for the wireless industry, and there will no doubt be more of this in the future until there’s no more spectral land left to farm. Carrier-aggregation cobbles together non-contiguous snippets of spectrum with the aggregate producing some additional bandwidth. The goal is to use beef up existing allocations to accommodate higher data rates and capacities. However, wireless applications will need much more than what can be obtained this way; a single channel will be at least 100 MHz and three or four times that, eventually.

All Millimeter-wave Spectrum Is Not the Same

Lost in some discussions is the fact that propagation and attenuation are not the same at every frequency in the millimeter-wave spectrum and other factors that make it challenging. Free space path loss, atmospheric absorption, scattering, and non-line-of-sight propagation are the most prominent problematic.

Free-space loss is the attenuation of electromagnetic energy between two points and the ability of the receiving antenna to capture the signal. At 28 GHz, this loss over a 10-km span is about 13 dB, including 10-dB of gain produced by the receiving antenna.  However, 10 dB of gain from a receiving antenna is too conservative at this frequency, as even smaller antennas can produce twice this amount. So, using a more representative 20 dB of gain reduces free space loss to that of 10 GHz.

Moving to 40 GHz where higher-gain antennas can produce 40 dB or more of gain, path loss would be about 104 dB. Note that none of these calculations include gain of a transmit antenna and over a distance much longer than would be typical in a 5G application. For example, adding 40 dB of transmit antenna gain over the same 10-km path at 40 GHz reduces path loss to 64 dB. If the distance is reduced to 0.3 km using a 40-dB-gain antenna, loss drop to 34 dB, and over a path of only 10 m loss would be about 5 dB, a manageable number.

The point of all this is to show that although path loss at millimeter-wave frequencies is comparatively immense, it can be mitigated by the large amounts of gain that is achievable at these frequencies. This discussion also does not include any benefits derived from the use of massive MIMO, beamforming, phased-array antennas with hundreds of elements, and other advantageous approaches. On the other hand, it also represents a line-of-sight path between two fixed objects, which obviously will not be the case for mobile devices. 

The next limitation is atmospheric attenuation, the elephant in the room at many—but not all—millimeter-wave frequencies, as shown in Figure 1. Signals are absorbed by molecules of oxygen, water vapor and other gaseous elements coinciding with the resonant frequencies of gas molecules, causing attenuation that varies dramatically with frequency.

At 28 GHz, for example, where Verizon is rolling out fixed wireless access (FWA) broadband service, attenuation from rain, fog, and moisture in the air is remarkably low, about 0.2 dB per km, and is even less at 39 GHz and still reasonable at 58 GHz, both frequencies to be used by 5G. Move just slightly higher, to 60 GHz, and attenuation doubles. Losses at all these frequencies is less at higher altitudes because moisture generally decreases with altitude. So, absorption at millimeter-wave frequencies is not universally “bad”, and extraordinarily high free path loss is less a factor at the very short distances for which it will of necessity be used.

Scattering, which is associated with atmospheric absorption, is certainly a factor as are bending (diffraction), shadowing, and other anomalies can also have an effect on the communications path. But they can be dealt with to varying degrees using the various techniques such as MIMO mentioned earlier.

Attenuation from foliage is a major concern as it can make communication difficult or impossible, limits range, and increases scattering. The severity of the effects of foliage is dependent on the operating environment, which makes it difficult to model or predict. Another issue is the inherent, and dramatic, reduction in signal strength caused by the line-of-sight propagation characteristics at these frequencies.

This challenge is one of the most studied by industry and academia throughout the world because without sufficient remedies it could derail millimeter-wave communications in many scenarios. Fortunately, many groups have proposed ways to mitigate the effects of non-line-of-sight paths using dynamic base station formation and other adaptive network approaches, which often include the use of sub-6 GHz channels when the higher frequencies become unavailable. In practice, all the approaches described here will be used throughout every network to produce a desirable result. 

Massive Available Spectrum

The major appeal of millimeter-wave frequencies is obviously the comparatively massive amount of spectrum available when compared with lower frequencies. For example, the entire spectrum ever used by wireless services up to 5.8 GHz easily fits within just the 60 GHz unlicensed band. The 28-GHz band offers 850 MHz of bandwidth, the 37 to 40 GHz band offers 3 GHz, and 7 GHz is available between 64 and 71 GHz.

The bands in which absorption is very high, such as 60 GHz, will still prove useful for 5G in certain scenarios such as for achieving very high data rates in areas densely populated with other radios. It is also beneficial when information security is paramount, as very narrow beams and very short distances combine to produce a low-probability-of-intercept communications path. These two characteristics are also what make operation at all millimeter-wave frequencies appealing, as they allow a very high level of frequency reuse, effectively expanding the capacity of base stations in a given area. Many companies are now testing new products at 60 GHz with specialized development  kits that can be used to develop a wide selection of product applications in the mm-wave unlicensed frequency band covering 60 GHz. Figure 2

Yes, Mobility Will Be Possible

There are some in the wireless industry who believe that mobility will not be achievable using millimeter wavelengths, although the extraordinary history of electronics technology in the last half century shows that view may be short-sighted. The first argument presumes that millimeter-wave communication is inherently impossible using hand-held devices because the user’s hand will dramatically reduce signal strength, negating any attempt at beamforming and MIMO. Other arguments point out that most trials only show a fixed-to-fixed scenario and are thus useless for determining the viability of mobility, and that producing mobility solutions will be too expensive.

All of these arguments are supportable, but the fact remains that the use of millimeter wavelengths is mandatory, not optional for 5G. That is, achieving the extraordinarily high data rates proposed for 5G mobile devices can only be achieved with a significant increase in bandwidth. This bandwidth is available only in the millimeter-wave region, and using every available technology at its disposal, the industry will provide solutions, and in some cases already has.

Take, for example, Qualcomm’s QTM052 millimeter-wave module that provides 800 MHz of bandwidth using frequencies within the 26.5 to 29.5 GHz, 27.5 to 28.35, and 37 and 40 GHz bands. In a device measuring 18 in. long x 5 mm high, the module integrates a 5G NR transceiver, power management IC, silicon RF front-end, and an entire 24-element phased array antenna that performs beamforming, beam steering, and beam tracking.

When combined with the company’s Snapdragon X50 modem, the QTM052 is essentially a complete multi-band radio and antenna. To overcome blockage, the QTM052 and X50 can switch to other beams produced and received by up to four modules mounted in the corners of the phone. The first version of the device was 25% larger, so it’s possible it or its successor will be even smaller than the current one.

The issue of blocking has been studied by more than a dozen organizations, and it has been shown that when combined with the infrastructure required to support mobility (i.e., small cells), a 28-GHz-enabled smartphone can communicate reliably over short distances. The studies took into consideration dozens of different hand sizes, positions, and other factors.

The small cell “resources” use massive MIMO, beamforming, and phased-array antennas that can only be affordably produced using standard fabrication techniques. Major strides have been made in these areas, and phased arrays incorporating hundreds of elements have been demonstrated. They are complemented by intelligent beam searching and tracking algorithms that switch to the dominant beam path, among other things. There is no doubt that millimeter-wave mobility is the most difficult challenge and will take the most time to fully realize. It will also be expensive as millimeter wavelengths have inherent limitations that require a long list of technologies to overcome.

The use of millimeter wavelengths is not a panacea for everything 5G aims achieve. It won’t be the primary spectral region used by 5G, nor was never intended to be. It will rather be accompanied by frequencies below 6 GHz, and together they will enable the increases in capacity and speed, and reductions in latency that future fixed and mobile applications will require. Probably the important factor to keep in mind is that the millimeter wavelengths are a new frontier. Until now, terrestrial millimeter-wave communications was limited to point-to-point links and a few other obscure applications. So, no one should be surprised if it takes a decade or more to realize the potential of this new resource.

(981)