Google, Microsoft, Comcast Say Verizon’s New Cellular Tech Could Wreck Wi-Fi
bloomberg.com
bloomberg.com
In a nutshell; the LTE transmission cycles on and off on a regular cadence without sensing if the channel is clear. This tramples over the 802.11 frames and results in higher utilization of the medium than that which is claimed by Qualcomm
I guess the bad feeling is that the unlicensed bands would be used for a subscription service, whereas WiFi is controlled by the consumer (though in practice each ISP ends up using WiFi).
I worry that it could make the 5G band look like 2.4G looks now. I too like the idea of a piece of spectrum used mainly by consumers or small businesses and not saturated by the big commercial gorillas. If today they are desperate enough to consider running their precious services on the wild unregulated bands then sooner or later they will be desperate enough to squeeze every last bit out of them.
Another thing is that once they are allowed in and their bottom line starts to depend on WiFi bands, they will fight to the death to stay there forever and only grow.
If they aren't, then -AIUI- 5GHz (and 40GHz) gear commonly used by WISPs is highly directional and requires somewhat careful alignment in order to work. (I've heard the term "pencil beam" thrown around to describe the antenna pattern.)
It surely should not be allowed until it's significantly improved.
Which is not to say you may not have had a bad experience with 5 GHz routers/access points -- there are plenty of commercial APs that are frankly terribly designed.
Although I believe all channels are high-power channels now (as of late last year or early this year?), switching to one of the old high-power-allowed channels can help. 149 is recommended.
I have yet to see any scientific paper that says 5GHz penetrates structural elements worse than 2.4GHz. Rain, yes. Walls, no.
If they are, my Linux systems aren't yet aware of the change.
They make the claim that the only ranges with a 30dBm transmit power are 2402 - 2472MHz and 5735 - 5835MHz. All other WiFi bands are 23dBm. (Though it does seem like the transmit power in the 5.49 - 5.73 GHz range got increased from 17 to 23dBm somewhat recently.)
> I have yet to see any scientific paper that says 5GHz penetrates structural elements worse than 2.4GHz.
I can't point you to any studies, but -anecdotally- I've found that (at the same transmit power, and all other things being equal) a 5GHz link provides a lower SNR than a 2.4GHz link. Before I switched to an AP with better antenna and radio, I found that I would not-infrequently permanently lose the 5GHz connection in my bathroom [0][1], but the 2.4GHz connection could be reliably established in the bathroom and remained solid.
Even after I've switched APs, I get a substantially better signal [2] from the 2.4GHz connection than the 5GHz one.
[0] My bathroom is tile covered, filled with pipes, and the furthest point in the apartment from my AP.
[1] "Permanently" as in "once I left the bathroom I could reestablish the 5GHz connection".
[2] Obviously, a better signal doesn't guarantee a faster connection. At my site, 2.4GHz is so overcrowded as to be nearly useless.
2.4GHz is to be avoided if at all possible.
Also, measuring SNR is interesting, but you should really run a speed test of some sort. A team I work with wrote this to let access points speedtest clients without any code on the client:
https://gfiber.googlesource.com/vendor/google/platform/+/mas...
(Only really tested with ath9k and ath10k. Could possibly work with other chips.)
Of course, there are many variables on translating link quality to actual throughput. Rate control algorithms are essential, and often wrong. For example, consider sending at 10x the speed and losing 2x the data. Most rate control algorithms will avoid that scenario, even though it's actually faster and conserves valuable airtime.
This is good reading on that subject: https://wireless.wiki.kernel.org/en/developers/documentation...
No doubt. Next time I find myself in something similar to my apartment, but in the middle of nowhere so's the three or four-dozen 2.4GHz APs that surround me don't confound the results, I'll do just that. :)
FWIW, the SNR measurements were taken with a UBNT SR-71E (802.11agn ath9k-driven) client that Linux reports as an AR928X.
I'll make a note to look at that wifiblaster program later this week. That could be pretty useful. Unrelated to that, how's the ath10k driver looking? Does it work most of the time for most supported cards these days, or does it still come with a boatload of caveats?
That said, now that the drivers are working, it's old technology. 4x4ac is all the rage now, and ath10k is 3x3.
Or use http://www.amazon.com/YSHIELD-EMF-Shielding-Paint-HSF54/dp/B... ;)
Why not?
It's too cheap to get good, enterprise-grade equipment for home usage that will have no problems and allow you to turn all of the knobs.
Support: On your hardware, do x.
Me: Did it already. It reported "Y".
Support: Please do it again, following these scripted steps [what don't match my actual hardware]...
[I pretend to follow along.]
Me: Done. It reported "Y".
Support: Oh, well then...
Because the hardware is often inferior to what you can pick up for ~$200 at a tech store, not to mention the lack of control you get with an ISP AP.
"lack of control": The Comcast provided hardware gives me an IP via DHCP and I'm able to disable all those useless firewall features. So which features am I missing and why should I care about them?
I can't use an AP that does not allow me to fine-tune QoS and just tinker in general if packets are not flowing in the way that I desire.
(not that my internet connection is always great - it isn't - but that lies with my internet provider, who I have no choice over, so...)
I think there are a lot of smart people on HN, but that doesn't mean they're all interested in all the same things. Tweaking wifi settings just seems like a waste of time to me, and boring to top it off. My interaction with my AP stopped at setting the ssid and passphrase.
Even if you aren't going to be hacking on your router much, it still definitely pays to ensure you're using hackable hardware.
This seems to contradict itself - how does it pay if you're not going to be hacking?
The commercial vendors do an absolutely horrible job of supporting or maintaining their products and they tend to get a lot wrong with the initial software release. The OpenWRT community does a great job of putting out a solid product that works and has sensible defaults, but you can only get the full benefit of their work if you buy hardware that is open to their hacking.
- fast enough to run Netflix and have a work VPN going at the same time
- fast enough to run Netflix, download something on Steam, and have a work VPN going at the same time
- fast enough to run Netflix, download something on Steam, have Dropbox sync multiple GB to a new computer, and have a work VPN going at the same time
But times have changed: On my old ADSL line QoS made a huge difference due to congestion (and the large buffer sizes). SSH was nearly unusable while uploading a file. My current Comcast line is so fast that I don't experience any issues due to congestion - so how would I benefit from QoS?
Another difference is that in the past consumer-level hardware was just unreliable: I remember an old Netgear router that just crashed if there were too many concurrent TCP connections, as its NAT table would overflow. I haven't seen those issues in a long time now.
Are you sure you don't have a few hundred milliseconds of bufferbloat in your cable modem, as is the case for almost all modems out there?
This argument has a place here, but it has a cringe effect each time I see it. I find it a dangerous one if applied to other circumstances.
http://paste.click/QEYrXr (Note the uid/gid)
Maybe wallpapers are too much pita, but if it came preinstalled in the walls and floors? Why not?
* beacons are sent at a low "common" rate, perhaps 6mbps
* beacons still include the preamble and the DIFS before that
* 802.11n/ac get high bandwidth from aggregating multiple frames
So all those trivially small beacons take little bandwidth but a significant amount of airtime (which is a significant amount of potential bandwidth for faster clients with multiple frames to send in one aggregate)https://en.wikipedia.org/wiki/Distributed_coordination_funct...
Here [1] somebody collected measurements of time eaten by beacons at few configurations. Not a complete disaster, but still somewhat significant. For example, at the place where I am now, I'm receiving 67 beacons per second (all at 1Mbps), which, according to those calculation, wastes 17% of airtime.
[1] http://wifinigel.blogspot.com/2013/08/its-well-known-rule-of...
By just counting the literal airtime of the beacons, I think it underestimates the effects a bit, because it doesn't account for the contention of the remaining air time, which would be reflected in increased collisions and small delays which (sorry to be hand-wavy again) can add up. I think if he ran some application-level tests at the same time (perhaps iperf, perhaps something more sophisticated) he would see a bigger impact to "good-put".
There's also a wallpaper (and even paint) version of this which isn't terrible expensive if you are living in an apartment building with exterior dry walls instead of concrete.
This is pretty much the "civilian" version of stuff government agencies use to RF proof undercover sites which could not be fitted with active jamming, it works extremely well.
I have one in the bedroom put it on because BT set up one of their metro wifi towers in line of sight of the window, it also blocks all the other wifi SSID's from outside.
Or look really close and hope you can see it.
LTE uses 2.5 but it's not that common 1900-1800 and 800-900 are still the most commonly used cellular bands.
I also plugged only a single window if you check which walls/windows leak the most and plug them it should allow cell reception through.
It would be extremely unwise to let them push LTE-U thru without lots of independent testing and analysis. Even then there is a case to be made for keeping some spectrum 'open' and unlicensed lest we see it all end up owned by a handful of corporations.
I think this is the important point. Let cell carriers in and they will subscribe as many people as necessary to fill the available bandwidth. After all, that's pretty much what they are paid for.
It's like road congestion - the amount of suffering depends on people's tolerance to endure suffering, not on the amount of available resource. There will always be more people willing to watch lolcat videos on their phones if they don't stutter too badly.
Keeping some spectrum aside from big operators allows at least small LANs to operate reasonably, while opening it to them may easily end up bringing suffering to everyone.
Wifi works. People will not be very happy if you single-handedly break all their wifi-enabled things.
Most people live in a state of ignorant bliss.
Your WiFi is slow ? Restart your computer, upgrade your operating system, clear your browser cache, move your router closer to you, buy a different modem etc. There is plenty of room for deniability that interference is the root cause.
Google is going to be pissed if their new wifi router doesn't work properly despite millions spent on testing, just because Verizon is screwing things up for everyone.
People will pile on, and the evidence that Verizon knows their technology can cause problems is already mounting. That they're pushing their products through without reviews or certifications could be seen as evidence of negligence or maliciousness.
What would be the cause of action? Why aren't the manufacturers of microwave ovens sued under the same cause of action?
If some product company (let's say, a router company) experienced some enormous increase in customer service and damage to brand image, I'd like to believe they'd eventually put quite a bit of money into finding out why they're losing money.
Could you imagine a world where a service rep has been instructed to ask whether a customer has any Verizon devices in the house? And to just give up, if they do, and blame Verizon? Would be terrible for Verizon's brand, pretty sure they're not ready (and probably don't want) to partition consumers who use internet-connected devices into pro-verizon and anti-verizon.
If you gradually deploy your non-solution while upselling them the cool new product, they won't even notice until it's too late.
I'm not sure about their zero-rating of audio streaming.
On the more universally reviled end of the spectrum, you have companies attempting to charge servers to pass traffic to their (also paying) customers.
On the nearly universally acceptable end of the spectrum, you have the widespread use of CDNs, or the Netflix and YouTube caching servers that they give ISPs for free.
Somewhere in the middle you have wireless ISPs offering free bandwidth to their users for use with popular services. Hard to get customers to see that as a bad thing, especially if they already use those services.
I was quite glad that the CRTC ruled against this practice in Canada. If telecoms want to run successful mobile television streaming businesses, they can start selling reasonable amounts of data with their plans.
People want bigger data plans more than they want Bell TV on their phones, so I think it got reasonable public support. Consumers groups widely applauded the ruling, at least.
First-party services fall in a different category; quite aside from net neutrality, that raises potential antitrust concerns.
But partnering with third-party services, especially if offering to do so with arbitrary third-party services, seems far less problematic.
Anti-trust law in the US is only relevant if the company has a monopoly.
That's exactly why net neutrality is important, and anti-trust law does not suffice. Because two competing ISP's both charging netflix to get the data you are already paying for to you, are not a monopoly.
(Notwithstanding that where I live Comcast is basically the only broadband option -- I still don't believe they have ever been legally determined to be a monopoly (yet?))
1. We made something awesome in labs (by bending/breaking some traditional rules)
2. It's X% faster than traditional connections/has all these properties, let's get to market ASAP
3. Oh, in the real world, this tech doesn't play nicely with other tech because of how we built it
4. Maybe we can just muscle it through?
If the scale were smaller (let's say, developers on some widely used open source standard/code), this would have been a simple case of "oops, that's a bit shit, better get back in the lab and try to figure out a way to not trample other broadcasts, but keep the speed boosts we were proud of".
Here is an analogy, its like a company wanting to sell paid access to a public beach.
"Let the market decide" doesn't work with commons. It's a tragedy, I tell you.
It sure was for Eric Garner…
Congestion-wise it will probably look like yet another WiFi AP, or zillion of them if it gets deployed sufficiently widely.
Cellular data is not another dumb pipe, it's data transmission grafted with duct tape onto GSM et al., which are ridiculously centralized and proprietary services. There are no open source implementations, no open source hardware, certainly no open source drivers. Even if there were, you would be at the mercy of whoever operates the base station.
EDIT: For all I know, the result used by Google is mostly simulation based (NOT TRUE).
Qualcomm did their due diligence on simulations AND lab trials to show different results. IMHO, Qualcomm clearly has a better arguments.
Note that the link provided in the top comment by rupellohn is old. The latest one, referred in the news link, is probably (there are multiple filings) this one: http://apps.fcc.gov/ecfs/document/view?id=60001331188
Full list of recent filing here: http://apps.fcc.gov/ecfs/proceeding/view;ECFSSESSION=Xp4PVvB...
TL;DR: 1. LTE-U has been proven to be a better neighbor to a WF AP than another WiFi AP. In other words, two WiFi APs in the same room perform worse than One WiFi AP and One LTE-U station.
2. Fairness: 2 WiFi APs should each share each have 5% airtime according to the standard, right? Wrong. Lab trials show one AP could take up to 80% +. Refer to http://apps.fcc.gov/ecfs/document/view?id=60001104452 "Wi-Fi/LTE-U Airtime Fairness". LTE-U always gives 50% Air time to its WiFi neighbor.
3. It is more of a political argument than a technological argument now. Note that at this point LTE-U spec does not violate the regulation on unlicensed spectrum.
DISCLAIMER: I don't work for Qualcomm. But it is easy to see who has a better argument.
I find it ironic that you've been "following this technology", have quick access to Qualcomm filing links, are happy to fill in details, but couldn't be bothered to even skim Google's filing.
The Jindal and Breslin study is well written and understandable. I'd suggest reading it since you seem invested in the topic.
I confused the filing by Google with other WiFi supporters (which uses simulation results). I have read through the all the filing months ago and apology for the bad memory. I agree that Jindal and Breslin study is solid and raised a valid concern. But they were all addressed in a response though.
Now I'd love to know how others think after reading Qualcomm's filing and not just take on one side of the story.
Section 3 of Qualcomm's filing did respond to many concerns in either simple analysis and/or lab tests. For example, the concern on the impact of rate control algorithm in WiFi.
The only thing I can think of is that they don't have enough high-frequency spectrum (1700/1800/2ghz+) which means lower theoretical data speeds.
They already violate their open/unlocked agreement for LTE and prevent other LTE devices from coming on their network and prevent theirs from going on other networks.
They are going to do whatever they can get away with.
* with extreme exceptions there are some just so they can claim they meet federal guidelines
Now you might be able to trick activate an LTE sim for verizon on a verizon phone and then get it to work on your phone but technically verizon doesn't allow it
One day there is going to be a really large lawsuit about this, specifically because it violates the spirit of LTE licensing if not the actual letter of the law.
What the do to prevent their phones being taken to other carriers is even more devious. To get around LTE portability requirements, they simply disable specific LTE bands like the ones for T-Mobile, etc. So yeah you can put a T-Mobile sim in there but the phone won't get any signal.
I don't necessarily think Google or (gods forbid) Comcast are above criticism, but in this case Verizon is wanting to sell the commons back to us. This is akin to Nestle saying water is not a human right, for a digital society.
Any time large bureaucracies run into a free market, they inevitably try to kill or regulate it (whether for the benefit of entrenched companies, or for state control purposes, or out of misguided nanny-state protectionism). There are few nations across the globe that haven't followed the same increasing regulation / control path with regards to the Internet, to one degree or another.
If the internet as we know it today ever dies, I'm 80% sure it will be due to private corporations screwing everyone to make an extra buck, not due to government over-regulation.
http://www.pcworld.com/article/259680/emerging_lte_direct_st...
https://www.qualcomm.com/media/documents/files/lte-direct-wh...
It gives the telcos a way to measure and control your access to their network and services so they can bill you for it. Wifi calling offloads everything but the actual phone call to a 3rd party (i.e. not Verizon) so Verizon can't keep their fingers in the pie.
So software tweaks, then. Yes?
In that world, noone wins.
[0] Other, cooler schemes require that WiFi radio manufacturers put more smarts into their radios. There are several 802.11 standards that would make WiFi devices more performant in the face of interference, -probably- remove the hassle of AP channel selection, and distribute relevant metadata about an AP to clients before they connect to aid the human in selection of the proper network and the machines in determining if the client possesses -say- credentials of some kind from a partner network that this WiFi network will honor. The benefit of adding stuff like this is pretty obvious to techies. However, because most computer users are largely non-technical, it's (apparently) impossible to make the business case to increase the unit cost of one's WiFi products by $20 or so to cover the R&D costs. :(
Verizon has huge chucks of spectrum licenses already to work with: (https://en.wikipedia.org/wiki/Verizon_Wireless#Radio_frequen...)
It's entirely Sprint's fault for not building up a reliable network (of course it's in the 2.5 GHz band) on what the FCC has given them.
Edit: Maps of the relative amount of spectrum for U.S. LTE providers is here (Sprint leads the way):
http://www.fiercewireless.com/special-reports/2015-how-much-...
A 'public' wifi hotspot is either run by, or connects to, an ISP. What's the difference there?
LTE-U will be used by the 'public' to access an ISP (that happens to be a cell phone company). By your standard, would all of the above be disallowed as they run through a company network to get to the Internet, or would they all be allowed as they are being used by the 'public'?
Sure, it gets quite complicated when you start talking about customer operated equipment that provides commercial service (xfinity is a good example) but perhaps the critical test would be "is the transmitting equipment fully managed and controlled by the individual?"
Or to phase that another way: There is a lot of spectrum. But very little spectrum that is suitable for mobile phone usage. When a resource is finite and valuable, it gets expensive.
If you want a "real" solution that would allow startups to play in the ballpit? Pass new national laws which split the cellular companies into two organisations, one that offers consumer services, and another which exists to maintain/install/upgrade the physical cellular network (and who then resells access).
It actually makes little sense to have competing and exclusive cellular infrastructure. The infrastructure should just be like a public resource, like a road, and the cellphone companies should be like shipping companies, utilising a public resource (the road) to sell a service (shipping stuff).
I'd be all for the government using eminent domain to seize every single cellphone tower in the country, and then reselling access to them to every cell phone company equally. But it isn't very capitalist...
Yes, that does mean that you end up with mega-corps providing your cellular service. But the same is true for say retail. Amazon will eat the whole industry. Physics as well as economics strongly favor mega-corps. Startups have advantages, but specific and narrow ones.
[1] See https://www.scribd.com/doc/292530059/Spectrum-Allocation-Loc...
This is similar to the approach that some countries used when privatizing railroad companies: Instead of creating a new private company that's effectively a monopoly, they created one company that provides the infrastructure and another company that runs the trains. This allowed new private companies to enter the market, as they were allowed to use the resources under the same conditions as the former monopoly.
I personally would be less than pleased to have to pay extra on an ongoing basis just to have a wifi router in my home, or a wifi radio on my laptop. Extending private property rights into what is currently a commons would seem to imply that though.
Here in New Zealand we have three providers, but we have national number portability. This means I can switch networks with no fees (unless I have a contract with a termination clause) and no number downtime. It helps competition a lot.
Number portability is already possible here in the US, and so long as I am not on a contract I can switch without fees.
Sure, this will even the playing field for coverage, but it won't even it for anything else (customer services, device financing, cost, et al). Plus actual network costs have little relation to what companies can charge, since all cellular companies over-sell their network.
I'm sure with this there will be companies that offer terrible customer service, and price their offering dirt cheap. That's fine. But most consumers will want to find a happy medium between the quality of service Vs. the cost.
That was the shape of the cellular networks in the UK in the early days of mobile telephony, IIRC.
There would suddenly appear to be huge amounts of available spectrum if the authorities would just yank bands from the most useless legacy allocations and auction them off.
I also suspect that the users of that spectrum (save for the military which has a huge amount of lightly used spectrum) might disagree with you as to how useless their technology is, I mean would you take the spectrum currently held for ham radio operators and sell it off?
Navigation (GPS) is pretty spectrum efficient, so I'm not really sure what you're getting at there.
All the legacy nonsense needs to be ruthlessly eliminated. People claim frequency auctions are efficient but that's clearly not the case if you are only auctioning .1% of whats useful and allocate the remainder to waste on a free cost basis.
Much of what's in that band is allocated for point to point microwave and satellite, I'd also caution you, one mans 'wasted spectrum' is another 'vital important service' both of which can be technically correct at the same time.
As you mention police, I'll also point out public service frequencies (Part 90) recently went to 12.5kHz bandwidth (half what hams use) and there is discussion of going even further.
The really legacy allocations like broadcast TV, AM radio, and radionavigation are in spectra that Johnny Startup probably isn't interested in. Data rates are pretty slow at 30kHz.
This is totally true! And it was a "junk" band to begin with. What if they allocated another unlicensed band that had room for 3-4x as many exclusive 802.11 channels as the current one?
Not sure which side I am on. I don't like Verizon, but I don't really support heavy regulation.
Truly unregulated just mean that everyone is entitled to screw everyone else, that's hardly anyone's intention.
The FCC still maintains ultimate regulatory oversight. And if LTE-U does cause issues with WiFi then it seems like something that makes sense to be regulated appropriately.
I'll tell you! (It's clear to me what side you're on.) You do really support heavy regulation, and you don't think some guy with a cheap and crappy device in the vicinity should be able to disrupt your cell phone reception, FM radio, the GPS wherever you are trying to figure out your location, or, yes, wifi. Likewise you don't want to have to drive from California to Boston because it's too dangerous to get in an airplane with all the unregulated recreational drones in the airspace. You want the radio spectrum, class A airspace, vehicle and factory emissions, and a bunch of other stuff to be heavily regulated. (What these things have in common I'll leave as an exercise.)
EDIT: this is at -1 but it's true. I'm leaving it. I guarantee that the guy is for heavy regulation and doesn't want me disrupting his cell phone, GPS, or wifi due to crappy devices that do not even give me any benefit but just ignore regulation / or there isn't any. he enjoys the benefits and is for them. I guarantee it.