We knew it was coming and the wireless industry is already bantering about the term 5G. Nobody knows exactly what it is going to be, but the consensus is that it’s going to be fast. The South Koreans are devoting $1.5 billion in research to develop the next generation wireless. And there are vendors like Samsung who are already starting to claim that the upgrades in their labs today are 5G.
And of course, all of this is hype. There is still not any broadband anywhere that complies to the original 4G specifications. This all got out of hand when the marketing groups started to tout 3G performance for networks that were not yet at the 3G specs. And then came 3.5 G and 4G, and now I guess 5G.
But let’s look at the one claim that it seems 5G is going to have, which is blistering fast speeds, perhaps up to 1 gigabit per second. What would it really take to provide a 1 gigabit cell phone data link? The answers can all be derived by looking at the basic physics of the spectrum.
Probably the first characteristic is going to be proximity to the transmitter. When you look at spectrum between 3 GHz and 6 GHz, the likely candidates for US deployment, then the math tells you that it’s going to be hard to send a 1 gigabit signal very far, maybe 150 feet from the transmitter. After that the signal is still fast, but the speeds quickly drop with distance. Unless we are going to place a mini cell site in every home and on every floor of a business it is not very likely that we are going to get people close enough to transmitters to achieve gigabit speeds.
It certainly is possible to generate speeds that fast at the transmitter. But such a network would need fiber everywhere to feed cell phones. A network with fiber that dense probably wouldn’t even need to be cellular and could handle nearby phones using WiFi.
We certainly need new antenna technologies and those things are being worked on in labs. I’ve written previous blog posts about the various breakthroughs in antenna technology such as with very bit arrays using large number of MIMO antennas. I think we can believe that antennas will get better with more research.
We need better processors and chips. A chip capable of receiving and processing a gigabit of data is going to be an energy hog in terms of the power available in a cell phone. Such chips are already here but they are deployed with bigger devices with enough power to run them. So we are going to need the next generation chip sets that will require less energy and that will generate less heat before any cell phone can actually use a gigabit of data.
We need carriers willing to supply that much data. Let’s face it, cellular networks are designed to provide something okay to many rather than something great to a few. Perhaps making cell sites smaller would help alleviate this issue, but it is a real one. If somebody is really dragging a gigabit out of a cell site there is not a whole lot left for anybody else. And this would require an increase the backhaul to cell sites to 100 GB or even terabit speeds if 1 GB phones became the norm.
Finally, we need a new FCC. Because the way that spectrum is divvied up in the US makes these kinds of speeds nearly impossible. Gigabit speeds would be easily achievable today if there were some giant swaths of bandwidth available. But our bandwidth is split into little discrete pieces and most of those pieces are further divided into channels. This makes it really hard to cobble together a big consistent bandwidth delivery system. We tend to think of wireless as a big pipe in the same manner than a fiber is. But it’s really a whole lot of discrete little signals that somebody has to join together to get a huge throughput.