Is Millimeter Wave Just Another Band for 5G?
The short answer is — absolutely NOT. Want to understand the reasons why not, and how solving the millimeter Wave (mmWave) challenges will be the determining factor between the ultimate winners and losers of the 5G race? Then read on.
mmWave spectrum is getting a great deal of media attention lately. Qualcomm announced world’s first 5G mmW antenna modules, before that they had announced thier first 5G small cell SoC (System on a Chip) supporting mmWave as well as sub-6GHz bands. Facebook announced the trials of its Terragraph network, which uses 60 GHz mmWave spectrum with 802.11ay Wi-Fi. Verizon and AT&T have publically stated that their first 5G systems will utilize 28 GHz mmWave bands. So, what is the significance of these bands and why are they different than others Let’s take a closer look.
Spectrum is the lifeblood of wireless networks. The demand for spectrum is insatiable, from the mobile voice eras of 1G and 2G to the data eras of 3G and 4G. Introduction of new generations of wireless technologies forces the industry to demand new spectrum. For regulators and governments, spectrum becomes the tool to entice the industry to become trailblazers in the new technologies. Besides, governments reap handsome financial benefits in the process. For example, the recent 600 MHz spectrum auction yielded $19.8 Billion to the US government.
The spectrum bands targeted for 5G can be largely divided into two groups: Sub 6 GHz and mmWave. The former includes all the bands used for cellular so far as well as a few new ones. The typical bands being included in this group are 600/700 MHz, 3.3–3.7GHz, 4.4–4.9 GHz and beyond. Sooner or later, much of the spectrum currently being used for 2G, 3G and even 4G will be refarmed for 5G. The mmWave band is defined as spectrum between ~30 GHz and 300 GHz. This band opens up a new realm of opportunities, and along with it, a whole slew of challenges.
The wireless industry is not new to challenges. Remember, many had claimed that CDMA defies physics; but it ultimately became the foundation of 3G. Many naysayers had ridiculed LTE (Long Term Evolution) technology by calling it Late To Evolve. However, LTE became the basis for 4G and is responsible for its glorious run. The challenges of mmWave can be divided into two categories: 1) Network coverage issues of this band; 2) Challenges with supporting such high-bands in portable devices such as smartphones.
The mmWaves being higher on the spectrum scale, their coverage footprint is very small. The typical cell site radius is 10s of meters vs. 100s to 1000s of meters for traditional bands. Additionally, mmWaves don’t penetrate through obstructions such as walls, vegetation etc. This means it would be impossible to provide indoor coverage with outdoor sites. And also that Line of Sight (LoS) between the device and cell-site becomes important. The mmWaves behave like light rays, blocked by obstacles but reflected off of surfaces. For example, you get the best light when there are no obstacles between you and the lamp, else, you will be in a shadow, getting light bounced off from other surfaces. So, what all this means is that the traditional approach of overlaying new spectrum on the existing site grid, which operators used during 3G to 4G transition won’t work for mmWave based 5G. A dedicated network design is a must. The mmWave network has to be designed for capacity, while relying on Sub 6 GHz band or Gigabit 4G/LTE network for seamless coverage. On the other hand, the benefit of smaller coverage of mmWave cells affords the option of deploying them much more densely, offering extremely high capacity. Because of all of this, the economics of mmWave network is far too different than traditional ones. Suffice to say, treating mmWave as another band is not the right approach.
Now, let’s look at the other challenge: supporting mmWave in a smartphone form factor. It is no exaggeration to say that modern mobile devices are one of the most complex technology marvels ever invented. Yet surprisingly, they are incredibly easy to use. This is made possible not only by the processor and modem chips but also the complex RF components and circuitry that connect them. This helps in sending and receiving the right signals from the network in the most and power efficient way. Supporting mmWave in devices poses a fundamentally different and more difficult challenge than supporting sub 6 GHz bands. The mmWave bands experience much higher losses and behave differently when going though cables, filters and other RF components. This means, unlike sub 6 GHz, the RF transceivers and the antenna elements have to be intelligently designed with minimal cables/connectors between them. Also, to make mmWave to work, it requires numerous antennas: 8 to 32 per device. All of this limits the location of the antennas on the device, which obviously significantly impacts the form factor as well as performance. Let’s look at one simple example of the challenges involved. Now remember mmWaves don’t go through obstacles? When holding the device, hands and others body parts become obstacles and block the signal, and you have to design mechanisms to overcome this adversity, say by putting distributing antennas around the device not just at one location. Just more proof that treating mmWave as another band does not work.
It seems deploying mmWave is a difficult proposition. So, why consider it at all? Well, wireless engineers, enjoy solving such difficult challenges. Also it helps that the rewards of solving mmWave puzzle are incredibly valuable. While available bandwidths in the lower bands range from in 10s of MHz up to couple of 100s of MHz, mmWave potentially offers 10s of GHz, i.e. more than 100-times! Imagine trying to ration drinking water in terms of bottles and cases; mmWave would be analogous to bringing in a giant water truck. So mmWave is the future. Looking at the data demand trends, the wireless industry can use all that bandwidth and then some!
Enough with the problems, what are the solutions? I must say, the mmWave train is just getting started. There are many solutions being worked on, with the initial ones showing lot of promise. For example, from the network side, relying on small cells is an attractive proposition. AT&T just signed a contract with the city of San Jose to use the streetlamp posts, for their small cell deployments. which is a very clever approach. The FCC recently announced easing of building restrictions on small cell deployments, which is also a very good development. Some operators such as Verizon are initially targeting fixed broadband use cases, which is a great way to start off with mmWave, and there are many others.
On the device side (as well as network), using large numbers of antennas (called as antenna arrays) are a must. Some of the leading silicon providers have already demonstrated mmWave solutions in smaller formfactors as well as shown mmWave working in non-LoS and outdoor environments. There are already a handful of 60 GHz mmWave Wi-Fi (aka 60 GHz Wi-Fi) commercial products in the market, giving an early glimpse of the technology at work. All of that shows great progress and point towards a strong future for this frequency band.
So, it is worth repeating : mmWave is not just another band. It needs lots of focus and innovation to make it a reality. But the rewards are worth all the trouble. The initial solutions are promising. But it is very early in the process. It is likely that whomever solves this puzzle most effectively, will dominate the 5G market for years to come.