In my last post, I mentioned some of the drivers creating tomorrow’s WiFi and 5G radio environment. Some notable ones are:
- Antenna diversity
- Frequency agility
- Dynamic transmit power management
This time around let’s talk about how these techniques are being further refined for 5G and beyond.
As was writing this column, I thought it curious that the four techniques above have been used for decades and do not rely on any recent “hard” technology breakthroughs like computers, software, semiconductor miniaturization or optical communication.
For example, frequency agility was first put into widespread use to protect military communications in World War I, in which the German side used “automatic controllers which continuously and automatically vary wave lengths” in a way that was difficult for the British to predict. An early automated version of hopping between discrete frequencies was patented by actress Hedy Lamarr1 and composer George Antheil in 1942. George Antheil had earlier composed the score for the first “experimental” film La Ballet Mécanique, which, among other things, required 16 synchronized player pianos2. The Lamarr/Antheil patent proposed a paper piano roll to control the rate and sequence of frequency switching. For her part, Austrian native Hedwig Eva Maria Kiesler was well familiar with weapons advances in the 1930s. She was married for several years to Fritz Mandl, one of Europe’s biggest arms dealers, and had dined with Mussolini and Hitler before fleeing Europe in the late 1930s. Serendipity! Hedy and George are both in the Inventors Hall of Fame for inventing the frequency-hopping radio.
Another feature of our radio heritage is the notion of “junk spectrum”. At one time radio frequencies were treated like diamonds; a few rare ones had magic properties (good or bad) while the rest were not worth bothering with. Today we realize that “junk spectrum” meant “we are out of good ideas” and look astonishingly at the huge waste of transmission capacity in legacy systems such as analog television, taxi dispatch systems, and other pre-digital radios.
The 2.4 GHz band was once considered junk due to its absorption by moisture in the atmosphere (a feature exploited by microwave ovens), and yet today more data is carried in that band than by all the cellular network bands combined.
Today there is no such thing as junk spectrum. It’s all out there and it’s all useful for something. Like 2.4 GHz, the 60 GHz band was once considered unusable due to its short transmission range of 10 meters (33 feet) or so at the usual consumer low-wattage power ranges. Today, one application being considered is “room area networks”, where the same 60 GHz band can be reused from room-to-room without interference.
The ultra-high data throughput available at 60 GHz makes it attractive as an aggregating pipe in a series of exchanges between dispersed devices and a central processor … in other words … 5G!
So here we are in the new millennium. What once was junk is now valuable; what was once a constraint is now a feature. Yesterday’s “too short” transmission range is now “ideal for densification”. With 5G, *all* available frequencies are interesting and *all* will be exploited. New bands are opening up, like the 28 GHz auction just completed. Each of these frequency bands has its own feature set (atmospheric attenuation, incumbent users, natural interference, regulations), which will be incorporated into 5G network designs.
The essence of 5G is to “use all the available spectrum, whatever it is and wherever you happen to be”. A 5G network will be an amalgam of legacy infrastructure and new infrastructure. In some sense it will be cobbled together out of what we have left from our collective radio history. But in another very real sense 5G is a deliberate effort to squeeze the most out of the ether by intentional design, rather than consume spectrum as a side-effect of implementing useful applications like television broadcasting or taxi dispatch.
While 5G is in the consumer news today, the real 5G story is in carrier operations rather than consumer devices. Carriers need to exploit as much spectrum space as possible to fully support the increasing number of devices using radio, but they will not necessarily use new spectrum directly with those devices. That partly explains the continuing slow roll-out of 5G and even explains arguments about what 5G is: the big story is internal to the carriers, not affecting end users much.
That highlights another little-noted feature of future 5G. The techniques listed at the top (antenna diversity, frequency agility, densification, and dynamic power control) will allow high throughput, low latency, resilience etc, as advertised, but *not all at the same time*. A series of 60 GHz links will provide great throughput for movies, say, but latency will be too high for multiplayer online gaming. Small but high-priority data can be speeded through in channels set up for the purpose. And of course, costs will be allocated according to the spectrum’s “best and highest use”, to use the economist’s euphemism for “charging what market will bear”.
There are other 5G features that I have not mentioned. “Carrier aggregation”, for example, will allow slices of radio time to be allocated to different carriers, which is much different from today’s “one signal, one carrier” model The 5G specifications provide an orderly way to implement the slicing mechanisms and (of course) the payment methods. While interesting and useful, this feature is not much different from buying space regularly in a supermarket aisle or on a container ship.
Because 5G is not really technology problem, social and regulatory issues will dominate 5G deployment, rather than technology. Densification means lots of 5G small cells; those small cells need to be mounted somewhere (either “of right” or by contract), they need power (and to pay for that power), and they need a backhaul path (and pay for that too). A lot of jockeying remains to be done between municipalities (who may control the real estate), electrical utilities (who may also control the real estate and need to get paid for power), telecom carriers (who claim to be much too poor to pay for the resources required), and the FCC (which is required to strike a balance despite everyone’s thumbs on the scale).