Back in December 2023 a number of interesting decisions were made at the World Radio Conference (WRC) in Dubai. One of the most important bands for which a number of decisions were made was the 6 GHz band. Today, spectrum in this band is already used for Wi-Fi, and it is pretty much the only remaining additional spectrum for cellular networks in the next 10 years. The decisions reached for this band make it look like all players took a big gamble.
From an Internet access point of view, spectrum in the 6 GHz band is of prime interest for two camps: On the one hand, there’s the Wi-Fi camp and on the other, the cellular network operators. So, before I continue with my thoughts on the 6 GHz band, let’s have a look at the status quo and why both camps where fighting to get as much spectrum in this band as possible.
Let’s start with Wi-Fi. Two decades ago, Wi-Fi first started in the 2.4 GHz band. At the time, the overall bandwidth of 80 MHz was a lot. However, Internet speeds grew quickly and the band soon became crowded. The next step for Wi-Fi was the 5 GHz band, in which, depending on the world region, around 560 MHz is available. The band is not without its pitfalls, as there are other users such as weather radar. Wi-Fi devices have to look out for such signals and change channels to avoid interference. If other users are found, networks have to fall back to the lower part of the band. Unfortunately, transmit power in this part of the band has to be lower than in other parts of the band, which limits transmission speeds. In practice today, the 5 GHz band still works very well, perhaps helped by the fact that the range of 5 GHz Wi-Fi networks is significantly lower than Wi-Fi networks that use the 2.4 GHz band. Personally, I live in a big city in Germany and have no problems with interference from other Wi-Fi networks in the 5 GHz band. With my current equipment, I use a 160 MHz channel with a resulting data rate of just about 1 Gbit/s. In areas of very high demand, such as conference areas, university buildings, airports, etc., Wi-Fi can also still easily cope with the demand, and there is still a lot of room to further increase capacity by deploying more Wi-Fi access points and reducing their transmission power.
Now over to cellular: Here, the story is very similar. 30 years ago, cellular telephony became a mass market phenomena. At the time, around 50 MHz of spectrum was used for the service in the 900 MHz frequency band. As demand grew, more base stations were deployed and the 1.8 GHz band with a total bandwidth of 150 MHz was added. For 3G, the 2.1 GHz band was put on top and for LTE, a myriad of additional bands were added to keep up with the rising demand. Base station densities in cities grew as well, and we are at a point today where it is close to impossible to add further sites, both from an availability and financial point of view. For 5G, a significant amount of spectrum was added around 3.5 GHz. In Europe, 500 MHz was made available. This was significantly more spectrum in a single band then anything that was seen before. In addition, it was decided to use Time Division Duplex (TDD) to separate uplink and downlink, which allowed to use more spectrum for the downlink than for the uplink. So instead of splitting the 500 MHz equally in uplink and downlink paths and requiring a duplex gap in the middle, almost three quarters of the spectrum is used for the downlink. Like in the Wi-Fi case discussed above, the range of 3.5 GHz channels is significantly shorter than that of other cellular frequency bands. Before the launch of 5G, there was a lot of skepticism in the industry whether the range of 3.5 GHz would be good enough to be used in practice. 5 years down the road, I would say that the 3.5 GHz band made a tremendous difference in urban areas. With new antennas and beamforming technologies, the range of the band, while still providing good data rates, is well beyond 1 km. This is far more than the typical inter-site distance of cellular base stations in urban areas. So at least as far as I’m concerned, the band has significantly surpassed my expectations.
All right, so that’s the story so far. If the amount of data in networks would not be on the rise, I’d say we are fine. But data volumes keep rising, both in fixed line networks, for which Wi-Fi provides connectivity for the last few meters, and in cellular networks that mostly provide connectivity for outdoor and mobile use. So I think it is fair to assume that both technologies will require more spectrum in the future to keep up with rising demand. Which brings me back to what was decided at WRC 2023:
When it comes to radio communication, the world is divided in three regions: Europe and Africa are region 1, the Americas form region 2, and Asia plus the Pacific are region 3. Since both Wi-Fi and cellular require additional spectrum in the future, I would have expected that both technologies will get a share of the 6 GHz spectrum in all regions. Surprisingly, that was not the case!
In Region 1, i.e. Europe and Africa, 400 MHz of spectrum was already previously assigned to Wi-Fi and staid in that camp. The remaining 600 MHz of spectrum went to the cellular camp. One can discuss if this is a good split or not, but give or take a couple of MHz, I think both Wi-Fi for fixed access and cellular for mobility and outdoor use have a perspective for the future.
In Region 2, i.e. the Americas, the complete 6 GHz band with well over 1 GHz of bandwidth went to the Wi-Fi camp. Cellular ended up with nothing! In other words, unless things will change in the future, there is no way for carriers in North- and South-America to grow in the spectrum area. But is it realistic that some of the 6 GHz band will be re-dedicated to cellular in the future? Personally, I don’t think so, as already today, Wi-Fi devices are on the market that can make full use of the band. Another option would be to allow concurrent use of cellular and Wi-Fi in the band. This has already been tried with LTE in the 5 GHz band in the US, but I don’t think it was very successful. The main problem: In order for sharing the band, cellular transmission power has to be as low as the power used by Wi-Fi access points. This makes cellular service over more than a few tens of meters impossible. This is of course not very interesting to provide a typical cellular service in cities, where base stations are typically separated by more than 500 meters.
So has Europe and Africa made the wiser choice? I hope so, but two significant uncertainties remain: First, adding yet another frequency band at a cell site requires additional transmission power. The question is if there is still enough headroom available before RF-limits are reached. Sure, one can always reduce transmission power by putting up additional base station sites. However, as I pointed out at the beginning of this post, that is a very difficult thing to do. And the second uncertainty, like the last time around, is, how far a 6 GHz channel can reach with a meaningful data rate. I was very positively surprised by the 3.5 GHz band, so perhaps we will see something equally positive for the 6 GHz band? On the other hand, I was very negatively surprised by the range and difficulties of the mmWave bands above 20 GHz.
So, long story short, it is by no means a foregone conclusion that splitting the 6 GHz band for Wi-Fi and cellular in ITU region 2 was the right choice. As far as I’m concerned, I would rather bet on both technologies succeeding in this band in the future than to only go forward with Wi-Fi. Time will tell, let’s review this again in 5 years. That would be 2029/2030…