With amusement I read a note by a cable advocate on spectrum. The note seems to be mostly hung up on the definition of what is low-, mid- and high-band spectrum. It completely ignores the history of the terms and the technical reasons for why these terms are used and how the spectrum bands behind these terms have changed over time. When I started in wireless in the 1990s, the delineation was clear. Cellular licenses at 850 MHz were low band, and PCS licenses in 1900 MHz were high band. Low band and high band were (and are still today) synonymous with easy to work with and hard to work with. PCS spectrum was hard to work with because it had trouble providing a good signal inside homes contrary to 850 MHz spectrum. These definitions changed over time as new spectrum was released; RF engineers figured out how to work with it and make it useful. Today we have engineers lauding the great in-building properties of C-Band spectrum which is in the 3.5 GHz band. Different times, evolving points of view.
We also have different perspectives dependent on different geographies as the way spectrum got allocated has been different. In the US, we divide spectrum into smaller geographical licenses, while many other countries allocate a nationwide license, like most European countries. Europe decided to have its “high-band” spectrum at 1800 MHz, the US at 1900 MHz… and both are now mid or low band depending on your preference. The US, where I reside, has a different history than the UK, where the cable advocate resides, and history and geography inevitably skew the perspective. We do not have an arbiter who by divine rights determines what is low-, mid-, or high-band no matter how some people want to be that arbiter and tell others they are wrong.
But, here are some of the things that do not change. At 6 GHz to 7 GHz, electromagnetic waves lose the ability to bend around objects and become line of sight transmissions. That’s why today (and I am sure this will change again and throw some people into a tizzy) we draw the line, depending on the country, for high band spectrum at 6 or 7 GHz. In the United States, for historical reasons, we draw the line for mid-band spectrum either at everything above AWS spectrum (2.1 GHz) or everything above the 2.5 GHz licenses up to 6 GHz depending on your perspective. Both points of views are perfectly justified, no need to call the spectrum police.
The reason why Analysys Mason drew the line at 3 GHz is because the spectrum situation below 3 GHz is largely settled and little if any further spectrum of great consequences will come from there. Ten or twenty megahertz from broadcasters will not make a meaningful difference when it comes to capacity or speed. The global conversation about 5G spectrum allocations focuses almost entirely on this segment of spectrum above 3 GHz. All efforts to harmonize mobile spectrum at the ITU are focused on bands above 3 GHz.
It’s ironic the cable advocate would point to GSMA’s use of 1-7 GHz as mid-band. GSMA also called for countries to make 2 gigahertz of mid-band available for 5G. Following that logic, the U.S. ranks among the worst-positioned of studied countries in meeting future spectrum demands, ranking 11th in total shortfall counting all assignments between 1 and 7 GHz. As an aside, GSMA also notes “the 6 GHz band is an important tool in satisfying demand for mobile… it is difficult to meet demand…without it.”
Also, where one exactly draws the line (which is an irrelevant argument) makes little difference to the gap between the different countries. Any way you cut it, where you draw definitional lines or whether you include or exclude low-power or shared spectrum, the United States has allocated significantly less licensed spectrum in what is today the key spectrum frontier (mid band) where different countries allocate different amounts of spectrum for different purposes. Arguing about a 10 or 20 megahertz difference in the gap is lawyerly myopia: a gap of around 200 megahertz is meaningful. The same people who argue that there is no gap are the same people who turn around and state that the US has slower speeds than some countries… obliviously missing the causal relationship.
The reality is that the US has more need for licensed, full-power, exclusive use spectrum than other geographies. This is partly because of the popularity of fixed wireless access in the US, something that European carriers have mostly ignored as they do not want to cannibalize their own DSL and cable networks. For example, why would Deutsche Telekom or Vodafone Germany, two of the largest mobile providers in Germany, offer FWA when their home internet offers are predominantly DSL or cable, while at the same time have driven rates so low that their profit margin is below the cost of capital? In the United States, the mobile operators are in a huge fight to bring more FWA, which is cheaper and has happier customers than cable, to more people. Cable is understandably fighting to prevent any more spectrum being made available to wireless providers and instead have it allocated to unlicensed use that they can use for Wi-Fi. For the last six years, cable has been winning the spectrum war in DC as they have successfully lobbied the FCC for 1200 megahertz for unlicensed use in the 6 GHz band. The differences in spectrum allocations and the sizes of the pipelines between the United States and other countries just underlines this imbalance.
The fact is the US faces a mid-band spectrum shortfall compared to other countries regardless of how you define it, and the solution is establishing a spectrum pipeline and allocating more midband spectrum to licensed wireless.
