Next Generation of Wireless -- 5G -- Is All Hype (backchannel.com)

Many people have promised us that 5G will be here very soon. And it will be the best thing ever. To quote Lowell McAdam, the CEO of Verizon, 5G is "wireless fiber," and to quote SK Telecom, thanks to 5G we will soon be able to "transfer holograms" because the upcoming standard is "100 times faster" than our current communications system 4G LTE. But if we were to quote Science, the distant future isn't nearly as lofty as the one promised by executives. Backchannel explains: "5G" is a marketing term. There is no 5G standard -- yet. The International Telecommunications Union plans to have standards ready by 2020. So for the moment "5G" refers to a handful of different kinds of technologies that are predicted, but not guaranteed, to emerge at some point in the next 3 to 7 years. (3GPP, a carrier consortium that will be contributing to the ITU process, said last year that until an actual standard exists, '"5G' will remain a marketing & industry term that companies will use as they see fit." At least they're candid.) At the moment, advertising something as "5G" carries no greater significance than saying it's "blazing fast" or "next generation" -- nut because "5G" sounds technical, it's good for sales. We are a long way away from actual deployment. [...] Second, this "wireless fiber" will never happen unless we have... more fiber. Real fiber, in the form of fiber optic cables reaching businesses and homes. (This is the "last mile" problem; fiber already runs between cities.) It's just plain physics. In order to work, 99% of any "5G" wireless deployment will have to be fiber running very close to every home and business. The high-frequency spectrum the carriers are planning to use wobbles billions of times a second but travels incredibly short distances and gets interfered with easily. So it's great at carrying loads of information -- every wobble can be imprinted with data -- but can't go very far at all.

Comments

Re:They's right, probably

[hey!]

Of course we have the population density; it's just not uniformly distributed. Sure, the US on average is not densely populated, but the average American lives in a densely populated area, and many of those densely populated areas are merging into megalopolises.

If you take a super-densely populated, highly afflurent neighborhood, it doesn't matter whether that neighborhood is Azabu in Tokyo or the Upper East Side in New York. The staggering concentration of wealthy people is capable of supporting anything that is feasible and desirable. Tokyo is the richest city in the world, and New York is a close second. Of the twenty richest cities in the world, eleven are in the US.

No, there are only two barriers to the adoption of 5G in the US: (1) A political culture reluctant to adopt and promote technical standards and (2) the fact there is no such thing as 5G.

Just like wifi

[raymorris]

You've described "just like wifi", then asked if there's any reason people won't use it for mobile phone service, instead of using a wireless carrier.

Well, do people use wifi for their phone and not have a carrier? That's essentially the same thing as what you're proposing. Do people do it? Nope, most people use a carrier with their phone. Why? Well, why DO you have a carrier?

Sure, theoretically everyone could run their wifi with an open guest network, then use voip from their phone. Aside from security concerns, let's look at one obvious financial issue. You're running the hotspot, which you presumably have connected to your cable internet connection to connect to the rest of the world. Two of your neighbors like to stream HD Netflix all night, falling asleep with movies playing. That's helping run your internet usage close to the cap amd you need to upgrade your internet if you're going to have enough bandwidth for you and your neighbors, so you ask the neighbors to throw you a few bucks toward the bill each month. They each pay you $25/month, which you use to pay the upstream cable bill. So now your neighbors are paying you for the connection - congratulations, you're a carrier.

Suppose you DON'T connect your microcell to the internet. Suppose it's a pure mesh, where you hand the data off to the next neighbor, who hands it to the next, who hands it off to the next, on and on until it hopefully reaches the Netflix server. So that kinda works for a month or two - it takes a long time to make several hundred hops, and sometimes a site in the middle gets overwhelmed and drops calls but it kinda works. Then somebody says to you "rather than sending your traffic over hundreds of micro-routers and hoping it doesn't get dropped, you can conect to our cell which has a direct fiber optic connection to an all-fiber network, and it's just three hops (via fiber) to reach Netflix." The fiber-connected cells work a shitload better than jumping hundreds of wifi-like connections, of course, so you want to get on their fiber-connected cell. Only $25/month.

Mesh networks CAN work, much like two cans on a string can work. If you're in a small farming community in the middle of the desert with no cabling and no connections to the outside, talking to someone 20 miles away by being routed through a dozen neighbors is better than not being connected at all. But compared to connecting to a multi-gigabit fiber network running on redundant $25,000 Cisco routers that can reconverge around a dropped link in under a second? Not even close. On average Verizon spends $6 billion each year on network upgrades. Last year they spent $11 billion. They don't spend that kind of money on something that's no better than Linksys.

Re:penetration

[Ol Olsoc]

Yes, I know, let's get the jokes out of the way first.

Now, seriously, EHF is going to be seriously attenuated by windows and walls, being pure line of sight communication. I suppose the assumption is that people would use wifi for anything indoors? What if your windows don't face a local tower? Multiplicity of towers to avoid that situation? Increased infrastructure costs associated with this?

This doesn't sound likely in the near term.

Marketing forces have yet to defeat the laws of physics. Between the frquencies that would have to be used, and playing with Shannon's limit - which can be violated up to infinite data transmission with an infinite amount of power, 5G speed will almost certainly be relegated to the laboratory.

Radio frequencies do not act all the same. Some, like HF, can give worldwide transmissions, some times at power levels that make your cell phone look hugely overpowered, and frequency dependent on time of day and the sunspot cycle. At other times be completely dead becaus of solar activity. Low VHF is a crapshoot, often dead, sometimes opening up. None of that is useable for data. At VHF, it is starting to get interesting, but an effect know as tropospheric ducting will sometimes cause signals from far away to come in and interfere with the ones you want.

Now the VHF and UHF and above frequencies have adesireable effect. They usually are line of sight only, and they become even shorter in transmission/reception distance as they go up in frequency. The 2.4 GHz frequencies used for wireless routers are actually well suited for Wi-Fi. They attenuate quickly enough to allow for other people nearby to have their own wi-fi devices.

But eventually we get to the point that the RF is absorbed by things like water, leaves, bodies. And if the 5G is going to be at 24 GHZ, well good luck. As noted, you'll need to be darn close to the transmitter, you'll need to have a good bit of power on both ends. So it's fiber right up to a few inches away.

5G is almost as much hypefail as Broadband over Power Lines, which had so many failure points it was an exercise in suck.