In July 2014, the FCC released its Spectrum Frontiers plan, which allocated up to four large swaths of spectrum in the millimeter wave (mmWave) bands, above 20 GHz, for 5G. This spawned a bit of a land grab, with Verizon snapping up mmWave spectrum with the acquisitions of XO and Critical Path, and AT&T acquiring FiberTower’s assets. It was a windfall for these companies, sort of the tech equivalent of having held onto a house in a lousy neighborhood that suddenly gets hot. With these acquisitions, AT&T and Verizon now own close to 60% of the licensed mmWave spectrum. The FCC still retains about 1/3 of it, and plans 5G auctions at some point. Verizon and AT&T have been marching down the 5G road, testing fixed wireless access in several cities in the mmWave bands as one of the initial use cases.
But even though 5G had been heading in a mmWave, circa 2020 direction, mid-band spectrum, characterized as that below 6 GHz, is proving to be an important contender for 5G as well. T-Mobile was among the big winners in the 600 MHz auctions completed earlier this year, acquiring 31 MHz of nationwide spectrum. In August, the operator announced that it would deploy a ‘5G Ready’ network at 600 MHz, meaning that new equipment from Ericsson would be used that supports both LTE and 5G at that band. Also in August, the FCC opened an inquiry into new opportunities in the 3.7-4.2 GHz band, to be used for the “next generation of wireless services”. This effort is backed by Google and several wireless ISPs, who would want to use this spectrum for fixed wireless services. At the same time, T-Mobile and the CTIA are leading an effort to make the 3.5 GHz (CBRS) band more ‘5G friendly’ by lengthening the terms of the licenses and expanding the geographic service areas.
The upshot of this is that mid-band spectrum is emerging as a viable alternative for 5G. One can see the battle shaping up, especially if Sprint and T-Mobile merge, which is looking increasingly likely. Sprint/TMO’s main 5G play would be in their 600 MHz and 2.5 GHz spectrum, plus leveraging their holdings in other bands as well (it should be noted that TMO owns mmWave spectrum serving about 1/3 of the country, through MetroPCS). It’s not clear how active they would be in a future FCC mmWave spectrum auction.
This is setting up a pretty interesting marketing battle and debate over 5G. The mmWave bands offer a huge amount of spectrum, which would deliver orders of magnitude improvements in network speed, capacity, and latency. The tradeoff is that mmWave spectrum generally requires line-of-sight, can be affected by weather, and offers a small coverage radius. Providing service in these high spectrum bands will also require the deployment of large numbers of small cells—and we haven’t yet found the formula to be able to do this at scale, yet. There is also still quite a bit of work to be done to develop the beam-forming antennas and other technology required to deliver wireless services in the mmWave bands.
So, what does this mean for the 5G rollout? We will see services marketed as 5G, even using sub-6 GHz versions of 5G New Radio, starting in 2018. AT&T has already prepped us by launching ‘5G Evolution’ in a handful of markets. In reality, these 4.5G, or Gigabit LTE services will offer considerable improvements in download speeds and latency, which are certainly in the neighborhood of what has been envisioned for the early stages of 5G.
It’s also becoming clear that there will be different flavors of 5G. Gigabit LTE, and other services offered in the mid-bands, will look more like today’s cellular services, supporting broad coverage and mobility. Think of it as a base layer. Then, mmWave band networks will be built in denser urban areas and other targeted coverage deployments, where it makes the best economic sense and where the most subscribers can be reached. The map will look like ‘islands’ of 5G in a sea of LTE and LTE Advanced.
It will be well into the next decade before there is broad coverage of mmWave-based 5G, and there is still some question regarding the extent to which mobility can be supported in these bands or how good the coverage will be in buildings. But thinking about 5G in this way, and with this timeframe, provides a good runway for the technology to evolve. Consider that the average LTE speed is 4-5x what it was only five years ago. Apply that multiplier to 5G, as a base case, and things start getting interesting.
In the meantime, fasten your seatbelts for the upcoming marketing war between T-Mobile/Sprint and Verizon/AT&T, over 5G. With no official body really calling the shots over the definition of 5G, it will be up to the market to decide.