Slashdot Mirror
Wi-Fi Gets Multi-Gigabit, Multi-User Boost With Upgrades To 802.11ac (arstechnica.com)
The Wi-Fi Alliance has announced its certification program for IEEE 802.11ac Wave 2, a technology that has been around on the market for more than a year. Wave 2 can deliver up to 6.8Gbps and lets an access point interact with more than one device at a time. Wave 2 features MIMO (or MU-MIMO) which improves the MIMO technology that lets Wi-FI transmit over more than one stream through the air. Wave 2 standard utilizes channels up to 160MHz wide (up from 80MHz channels available with Wave 1). It also creates more spatial streams and uses spectrum more efficiently, the industry group said on Wednesday. Ars Technica adds:On top of MU-MIMO, wider channels, and more streams, the Wi-Fi Alliance says Wave 2 features now being certified bring "support for a greater number of available channels in 5GHz," a change that "makes more efficient use of available spectrum and reduces interference and congestion by minimizing the number of networks operating on overlapping channels." You may have already noticed routers supporting some of these features, since the specification details have been available for a few years. In fact, routers with MU-MIMO started appearing in July 2014, and you can find routers that use 160MHz channels. The certification program takes a while to catch up with real-world implementations, but it ensures compatibility between devices and may spur faster adoption by vendors. End-user devices such as phones, tablets, and laptops must also be updated to take advantage of new features such as MU-MIMO.
god i bet you're still on IFITL fiber.
I am now going to spend the rest of the day muttering moo-mimo under my breath....
My eyes reflect the stars and a smile lights up my face.
Meh, if you live in an urban setting the 50 assholes who live within wifi range and use a 40MHz channel width because they don't know what they are doing you'll get shit speed and connection anyway due to the interference.
That is true on 2.4GHz. AC uses 5GHz. More channels, less people.
Be or ben't
Actually if the standard garden hose in the US of 5/8" diameter represented 1.5Mb, assuming an equal rate of flow, I believe the 6.8Gb in 802.11ac Wave 2 would be represented by a pipe 60" in diameter.
Is there anybody knowledgeable that can explain how a a wi-fi channel with bandwidths in the tens of MHz can deliver 6.8 Gbps? Even with MIMO, I feel like that shouldn't be possible (unless maybe they are talking about the 60 GHz band). Did they change the modulation to some crazy scheme?
Fewer people. Less range (which is one reason there are fewer people in range). Still better: Gigabit-Ethernet.
i use 60MHz you insensitive clod
maybe you should buy a house close to civilization?
Thanks for the reminder that I screwed myself by getting an integrated cable modem/802.11ac router. Meaning the router is controlled by Comcast. Meaning the firmware cannot be upgraded. Ever.
I've abandoned my search for truth; now I'm just looking for some useful delusions.
The "tripping over the cable" problem?
I've abandoned my search for truth; now I'm just looking for some useful delusions.
6800 / 1.5 is 4533. .625 inch pipe. = 2833.33333333 inches expanded to ratio.
So that's actually a 236 foot wide pipe.
Build a death ray from an old microwave and disrupt their wifi exacting your revenge?
So what would people do if they knew what to do? Would they hold a planning/coordination meeting where each neighbour gets assigned a channel, or what?
6800 / 1.5 is 4533. .625 inch pipe. = 2833.33333333 inches expanded to ratio. So that's actually a 236 foot wide pipe.
Everyone else it talking about round pipes. Your pipe is 5/8 inch talk and 236 feet wide.
It's about to be true for 5GHz as well, with Wave 2. The "Extended 5 GHz channel support" feature they are touting is carefully worded to whitewash the fact that 160MHz channels ruin channel plans by reducing the number of available non-interfering channels down to (with new FCC regulations) 4, only one more than 2.5GHz. Of course, nobody will be bonding 320G channel groups, so at least there won't be that part of the problem, and the additional channels are certainly welcome for 80MHz purposes, bringing the number up from 5 to 9. And 4 channels is certainly topologically much more than 33% better than 3 channels. It is a big relief that wave 2 didn't have to roll in on a 2 channel plan.
Wave 1 did indeed encourage device makers to actually put the 5GHz antennas back in, and wave 2 certification would encourage them not to leave any channels out (a.k.a. the "crummy hardware is why you cannot use 144 in a BYOD channel plan even though most systems support it" problem) and you don't *have* to turn on 80MHz, so there is some benefit to wave 2. Just, the hazard that prosumer equipment will come configured out of the box to run 160 puts us in a 4 channel universe when you don't have the authority to control spectrum use in a premises. Fortunately the standard also has some better avoidance/graceful-downgrade paths, whose full functionality I certainly hope are *required* for certification.
Someone had to do it.
Is this the kind of thing where most wave 1 devices can be software upgraded to wave 2 devices? Or is it yet another case of tossing out the silicon?
I would guess the glowing vendor support for this on one of TFS links would lead me to believe this will require new hardware.
From an AP support perspective, it really is annoying to have so many active fucking client standards to support. All the gee-whiz latest features are marginal benefits when half the spectrum is used by brain damaged clients vomiting all over the spectrum using old standards.
GigE- perfect for the mobile crowd...
Well, that was a nice, easy solution, now wasn't it? I'm so glad you were able to come up with such a simple and obvious fix!
www.wavefront-av.com
When Grandma gets it, ti becomes My Meemaw's MU-MIMO.
www.wavefront-av.com
True, but keep in mind that what matters is not just bandwidth, but bandwidth*time. If you have 100Mhz of spectrum to use, and user A and user B each get 50Mhz, let's say this gives them each 50mb/s of bandwidth. But if we let them both use all 100Mhz, then they each get 100mb/s peak bandwidth, while still having ~50mb/s each if it's congested. So if each of them has a fixed amount of data that they want to transfer, they use the same bandwidth*time as before. It only becomes a problem if you have a bandwidth hog who will use any available bandwidth you give them.
You mean to tell me that all the Ethernet cable I just finished running through my house is pointless.
Dammit!
That's wrong. I came up with 233 degrees Kelvin. That works out to around 3.2 acres for you metric people, and 85.67 Euros for the USA.
So does this mean it will finally stream video as well as 100mb hard-wired Ethernet?
Strange things are afoot at the Circle-K.
6800 / 1.5 is 4533. .625 inch pipe. = 2833.33333333 inches expanded to ratio.
Actually a 4533 ratio of hose cross-section area is equivalent to sqrt(4533) = 67.3 of diameter ratio. Therefore the 6.8Gbps hose would be 67.3 * 0.625 =~ 42 inches wide. Not very far from GP's estimate.
And a range of 10 centimeters.
There were LANs before there was even internet access! Companies pay big money to upgrade their LANs (and WiLANs) even as their internet speeds don't increase.
If you've really got no use for it, no problem. Why you feel the need to announce it to a group of strangers is beyond my comprehension, but nevermind that. I can see millions of uses for it, even with slow internet speeds.
For the average person, you'll probably also see much better speeds when you're out on the fringe of the signal area, than you would with slower implementations, making it usable a greater distance from the router.
Plenty of people have DVRs, movie libraries, HDHomeRuns, and similar, where higher speeds would greatly help them view movies on their phones/tablets/rokus/etc.
Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
There is some benefit to contention issues from packets taking a smaller time. However, peak bandwidth is mostly worthless (as well as rarely achieved in RF) when put shoulder to shoulder with latency and packet loss. Which is why 11ac has fallback provisions for reducing the bandwidth, cutting slices out of channels to deal with interference. When you have a residential area where spectrum is not planned, or not enough staff/money to throw at really tuning things, contention mechanisms won't always work due to hidden nodes/APs Really the "160Mbps" is mostly a selling point for PHBs. Personally if they can get away with it, I can totally see network admins only turning on 160MHz channels near the CIO's office and the meeting rooms were they go the most, There are plenty of nice features in wave 2 which make it desireable to have (in about 5-10 years depending on the age of your client mix), 160MHz channels are not one of them.
Someone had to do it.
It does if the pressure is sufficiently low.
I am TheRaven on Soylent News
Faster transfers between my laptop and the NAS that hosts the backups are always appreciated. If I don't need to plug in a cable, then that means that my backups are going to happen more frequently.
I am TheRaven on Soylent News
So when are we going to get this: https://threatpost.com/ibm-unv...
I mean it's not like I've been waiting or asking for it for years: https://it.slashdot.org/commen...
https://mobile.slashdot.org/co...
Shared key WPA2 means that anyone who knows the shared key can decrypt other people's traffic if they managed to sniff the 4-way handshake messages:
https://mrncciew.com/2014/08/1...
http://www.howtogeek.com/20433...
It's true using WiFi means you still have to trust the entity providing it, but that's the same with a wired network or using an ISP.
To those who say "use VPNs" I'd say:
1) Defense in depth
2) that's a different layer - just because you can workaround a broken layer doesn't mean the broken layer isn't broken. The fact is the layer already has encryption but it has a broken implementation which can be improved.
At least use the 20Mhz channel width that allows for 3 non-interfering channels. But yes, a planning meeting would be good. Maybe coordinate to buy a gigabit fiber connection and share it.
Yes it's an anecdote! Were you expecting original research in a Slashdot comment?
I've found the only way I can use 2.4GHz where I live is to flood a band all day with netcat from /dev/random. Eventually the firmware in other devices moves away from that channel, or the owner manually reconfigures, leaving it free for me.
I haven't had to bother for years because 5GHz is relatively clear, but it's slowly starting to fill up for shitty ISP supplied hardware.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
With more than (two? three?) antennas there are some issues that keep you from synthesizing and extracting as many completely independent channels as there are antennas at the end with the fewer antennas, but it approaches that.
Also: If you've got one central site (like a hotspot or cell tower or coherent array of them) with a lot of antennas and a number of remote devices with only one or a few, you can do things like "steerable null" - computing waveforms that send signals to several remote sites arranged so that each "can't hear" the signals intended for the others. (You make the others maximally quiet, rather than his maximally loud, because the ones intended for others are "noise" and its the ratio of signal to noise that matters.) Your multi-antenna remotes can also get multiple re-uses of the bandwidth and your single-antenna remotes a signle use of the bandwidth - but the cancellation on the multi-antenna remotes isn't quite as good, so it's another case of you can't quite get to data rates equivalent to the ideal of M antennas at the central and a sum of M antennas at the remotes giving you M complete re-uses of the spectrum.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
Lost the start of that, trying to recover . Quick summary.
Imagine a flickering-LED optical transmitter on the side of one hill, and (instead of an optic fiber to a receiver) a telescope pointed at it with a photodiode at the corresponding point of the image. The Nyquist limit applies to the amount of data you can send that way (and forward-error-correction coding schemes exist that approach the maximum theoretical bitrate).
Now imagine a billboard on the hill, with an array of flickering-LED optical transmitters, and an array of photodiode receivers at the telescope image plane. The individual signals are flying through the same space and separated by their direction of flight. The Nyquist limit applies to each of these links, but if the optics are good, the aperture large enough, and the air is not dusty, foggy, or the signal so strong it starts ionizing things, the signals don't mix enough to raise each other's noise level. So with N signals you can get up to N times the data rate. (If the separation isn't perfect you get less than Nx because of some of the signals appearing as a noise in the reception of others.)
At any cross-section of the path from the billboard to the telescope, the light is the sum of the light from all of the transmitters. The direction of flight of each individual signal is represented by the way the phase of its contribution changes from point to point across this cross-section. The telescope samples a cross section and the optics "computes" the separation of the beams, bringing each to a focus on a different receiver.
With MIMO the transmitting station does something approximately like synthesizing the signals at several points on such a cross-section and pushing each out through a separate antenna located appropriately. Meanwhile the receiving station does something like sampling several points on the cross-section and computing the separation of the individual flying-slightly-different-directions signals.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
Oh, look at the chickenshit AC too afraid to even attach a pseudonym to his post. Fuck you.
www.wavefront-av.com