The Future of 802.11ac 125
CowboyRobot writes "The 802.11ac standard is expected to be ratified in 2013 and NetworkComputing has an interview with representatives of Cisco Systems and Aerohive Networks about what that will mean for everyone else. 'Out of the gate, the increases in performance over 11n will not be tremendously impressive. The second wave--which will require a hardware refresh--gets far more interesting... First-generation 802.11ac products will achieve up to 1.3 Gbps through the use of three spatial streams, 80-MHz-wide channels (double the largest 40 MHz channel width with 802.11n), and use of better hardware components that allow higher levels of modulation and encoding (up to 256-QAM). Whether we will actually see 802.11ac products capable of 6.9 Gbps is dependent on hardware enhancements on both the access point and client that are not certain.'"
Re: (Score:3)
for 802.11y hardware
The real speed will arrive with 802.11xxx - specially designed for broadcasting 3D porn ;-)
Re: (Score:1)
Hah, that's nothing. 802.11ac was developed by those trolls who keep posting stuff with my slashdot account!
-
A. C.
Re: (Score:1)
The real speed will arrive with 802.11xxx - specially designed for broadcasting 3D porn ;-)
Are you implying there are network protocols designed for other uses primarily?
Horse Hockey!!!!!! (Score:1)
Omnidirectional, not really. Directional? Hell yeah, I have Ubiquiti Powerbridges deployed and I'm pushing ~297 Mbps.
That's what the AP's dashboard reports. And tyhe manufacturer and its Apple-esque acolytes will argue that that's "radio" speed. But, run iperf , or some internet bandwidth tester and get back to us.
Actual TCP/IP bandwidth on your ~297 Mbps link is less than 100Mbps. Hell the PowerBridge only has a 10/100 Base-TX ethernet interface! How do you suppose they squeeze 297Mbps over 100Mbps? Yea. Horse hockey!
Re: (Score:2)
I don't know about the GP's Powerbridge, but I had a pair of Ubiquiti NanoStation Loco M5's providing bridge service between two of our offices that got real-world throughput of around 75 Mbps (Measured by transferring a 4GB file using Windows Explorer to another Windows box, and stopwatching the time). The offices were only a hundred yards apart, but I had the TX power cranked down to minimum.
I've measured similar on other wireless routers. Haven't tried on Gigabit connected routers. So on to the quest
Re:I'm a wire guy (Score:4, Insightful)
I've had just over 200 Mbit/s with 5 GHz 802.11n, sitting in the next room from the AP, with a wall between us.
And probably noone else using the 5 GHz band in your area - am I right? Because as soon as the signal / noise ratio decreases the high efficiency modulations stop working and you must live with much lower spectral effifciency - that is' much lower usable bandwidth for the same slice of the spectrum.
Unfortunately increasing the transmission power doesn't help a lot - after all, the neighbour wants high bandwidth as well leading to an arms race.
Re:I'm a wire guy (Score:5, Interesting)
Remember that 5Ghz has a much shorter range than 2.4Ghz so the problems with traffic congestion won't be as bad. If all of your neighbours switched to 5Ghz, you'd still see a noticeable real-world improvement over 2.4Ghz.
Re: (Score:3)
Re: (Score:2)
The problem is just too damned many people are jumping on the wireless bandwagon and even the ISPs have started handing out wireless routers
I remember my first experiences with wifi back in 2006. I recall that the DSL ISP started making it an option back then. G was seldom seen, and everyone was on 2.4ghz, except for the Wireless A guys ( the routers used to tout ABG big time, though WEP was still king and XP didn't quite support WPA out of the box).
A year later I started getting other people set up, buying a few routers and handing my older ones down as gifts. I eventually settled on a dual band one in 2009 and have been pretty satisfied. Howe
Re: (Score:2)
Re: (Score:3)
"I haven't seen more than low single digit MB/s over wireless LAN, even under line of sight conditions with hardly any interference."
You must be using shitty hardware. We're using ubiquiti hardware at my office and getting the expected speeds.
Re: (Score:2)
I've got a cheapass n router and I get 11 MB/s, which is pretty decent considering it's only got 100 Mb/s into it.
Re: (Score:2)
Re: (Score:2)
I have to ask: Is anyone getting even close to these advertised transfer rates in real world scenarios? I haven't seen more than low single digit MB/s over wireless LAN, even under line of sight conditions with hardly any interference.
Same problem here. At 25 feet, 100mbit wired trumps 802.11n significantly for streaming video. It's good for cellphones, laptops, and other toys but for moving anything other than tiny amounts of data it's insufficient for me. Of course, I've got everything running gigabit wired now, so this really won't even be slightly interesting to me until the "second wave". By that time 10GB wired or better will probably be running through my attic... so we'll see.
Re: (Score:2)
Yes, something is definitely wrong there. Even in our most crowded areas on campus our clients manage more than that on raw tx/rx rates -- of course, having that raw "wirespeed" doesn't mean you get that amount of data throughput when tens of other clients are also using the "wire".
Re: (Score:2)
I have to ask: Is anyone getting even close to these advertised transfer rates in real world scenarios? I haven't seen more than low single digit MB/s over wireless LAN, even under line of sight conditions with hardly any interference.
In 2010 or so I was sadly surprised that my backups (PC Hard drive - > wifi laptop ) hit some invisible barrier of around 1.5 to 2 MByte/s from Windows host to laptop (Ubuntu or Windows). That was despite connections being between 11 and 54Mbit (yeah, the speed negotiated was pretty unstable for some reason, AND overhead plus bit to byte conversions kills about 90% of the fat number we see on the box). PS: Two different rooms, with a drywall separating both machines. The total separation was less than 10
Not sure the big deal here... (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
So many variables though. I have one of those cheap Android HDMI computers [amazon.com] on my TV. When connected via 802.11G streams were somewhat intermittent, even with line of site to the AP. When I switched over to a wired connection [amazon.com] all the stuttering stopped, and even web content like Ustream improved.
I suspect a poorly designed antenna in the Android device, but it could just as easily been bad drivers, interference from neighbors, or another device using bandwidth on the wireless channel. Some other big differen
Re: (Score:2)
TV recordings from mythtv are about 4 Mbit/s too (standard def). There are significantly more interruptions on "11n" "Lite" gear (with just one antenna, claimed ~100Mbit/s) compared to Gbit wired. One killer is buffering when skipping back and forward. If it takes a second to buffer a second of video, that's very annoying.
Compressed video streaming isn't a great argument for anything beyond high-quality full speed "n" components though. Raw speed is great when moving big files around, if the CPU and hard dr
Re: (Score:2)
Re: (Score:2)
Well, it is designed for you to stream 3D porn at Blu-ray quality across your house.
Re:Not sure the big deal here... (Score:4, Insightful)
My somewhat extensive experience with mythtv and wireless is that you need speed to work around latency due to interference or random multipath or whatever it is that occasionally slows stuff down. If you've got 1000 mb to transfer over 1000 seconds then on average you only need average speed X. However if you need to transfer exactly 1 mb every second, or the picture breaks up, and you occasionally endure 9/10ths of a second interference/outages, then you need 10 times the average speed to deliver. Or a bigger buffer, which means a long spooling up delay.
A good IT analogy is its like the difference between batch processing and a realtime OS.
Or maybe a standard /. car analogy is something like if you've got a 200 mile range gas tank, it doesn't really matter where the gas station is as long as its less than 200 miles away when you have a full tank... but the instant that the closest open gas station is 201 miles away, you're all done. Maybe thats an awful analogy...
No wait I've got a better car analogy. My gas station can deliver something like 5 gallons per minute, which seems like gross overkill for my fuel injectors which barely burn 2 gallons per hour on the highway. However the key point is my fuel injectors do NOT use 2 gallons per hour, they really use a microscopic droplet 60 times per second. Or something like that. Too early in the morning...
Re: (Score:2)
At 55 mph your car gets 40 mpg, and at 70 mph your car gets 20 mpg, and you've got three gallons of fuel and you've got an appointment in one hour, seventy miles away, and you've only got two gallons of fuel. No, that sucks too.
How about, the highway can carry 100 people per unit of length at 70 mph without collisions or it can carry 200 peopler per unit of length at 35 mph... no, wait...
How about we just only make car analogies when they make sense :/
I know, I must be new here
Re: (Score:2)
That's a trick question. The temperature in Seattle is always 43F.
Re: (Score:2)
Ok, so lets say that you are driving across a bridge and while you are driving across it it starts to collapse behind you at a certain rate. If at any time you get below that certain rate the collapse catches up to you and you die.
Re: (Score:1)
Or you could get your girlfriend to run around naked.
Oh, wait, this is /.
Never mind and carry on.
Point to point links? (Score:2)
Re: (Score:2)
Better performance in higher density areas. If you're within wifi range of 50 residences, each of which has their own wired uplink, they still have to share the same wireless spectrum. More is better.
Not much of an improvement?!? (Score:1)
Am I missing something here? Using math "up to 1.3 Gbps" is more than 4x N's "up to 300Mbps" which was a huge jump over G's "up to 54Mbps" so, apparently you need to be 5.5x or faster to be classified a huge jump for cisco people, a measly 4.3x doesn't do it.
Congestion & old nets = little benefit (Score:5, Insightful)
802.11ac isn't out yet but I have little hope of it really helping. I live in an apartment building I can already see 50+ routers on 2.4 & 10+ ON 5GH.
I just don't see that much of a benefit unless the congestion avoidance is really better than 102.11n.
Re: (Score:2)
Re: (Score:2)
I was hoping the cable company would purchase the router and place on pole around my house
Just wait until you have trouble, call the ISP's tech support line and they tell you to go up the pole, unplug the router, wait ten seconds, then plug it in again.
Re: (Score:2)
Re: (Score:2)
Move to France, I actually have most of what you're asking for as I am on Free.fr.
My Freebox server v6 provides:
- ADSL connection
- 802.11n (2.4 & 5Ghz)
- a 250Gb NAS
- DECT for my wireless phones (& free calls to much of the world - to almost all places I call)
- 4 Gb ethernet ports
The Freebox player provides:
- television over ADSL or antenna
- recording emissions to the Freebox server
- playing videos from all DLNA servers on my LAN
- airplay from the Macs & iDevices (no video though)
- Digital connec
Re: (Score:3)
If you transmit at higher speed it takes less time to send a given volume of data, allowing everyone else more time for theirs. Compared to 802.11n which will use 50%+ of the available bandwidth for streaming HD video the new AC standard might only need 10%.
Of course all that is mitigated by a single 802.11g router being maxed out by someone doing a download.
Re: (Score:2)
Does 802.11ac slow down in the presence of older networks like 802.11g slows down in the presence of 802.11b clients? All you needed was one old client present on the network for all transfers to slow down by 50%.
Yeah, you could set the access point to be 802.11g only but even then the 802.11b would keep interfering with the access point...
Re: (Score:2)
Depends what you mean... The slow down was actually due to having to wait longer between the end of one transmission and starting another one to give 802.11b clients a chance to jump in. I don't know if it matters for 802.11ac or not.
Re: (Score:2)
The inter-frame spacing is mostly the same. It's tiny compared to contention window handling and actual frame duration. It's not really the main reason for drastic slowdowns in mixed networks.
The main issue in mixed networks is:
* having to enable RTS and CTS-to-self frame protection to interoperate with legacy stations that don't understand MCS rates, and
* just sheer length of non-aggregate frames (ie, 11abg frames, and 11n stations that aren't doing aggregation - eg if they're doing voice data that isn't b
Re: (Score:2)
Thanks for the info.The thing is, for many people like me in noisy environments, 11ac with it's use of larger frequency windows will give us even more contention. As nobody has replied to my question on it including better contention, it appears that 11ac's advantages over 11n for us will be slight to none.
Re: (Score:2)
Actually, 802.11g only does NOTHING to prevent this. It's not the 802.11b client on the network, it's just an 802.11b client on the same cha
Re: (Score:2)
But the download will either uses less time-share, or complete faster. Depending where the bottleneck is.
Re: (Score:2)
Not if 11ac access points slow down automatically when seeing non 11ac access points like 11g access points do when they see 11b traffic...
Re: (Score:2)
Yes and no, dot11ac requires 5GHZ radio support, and there are more channels available there -- and also consequently on a 5GHz network enterprises can pack APs more tightly without turning down the power level. So dot11h and other frequency conflict avoidance schemes should allow APs to automatically avoid each other. However, because of the 80MHz channel option, which will doubtless be turned on by just about everyone, this advantage is mitigated to half over a 40MHz dot11n network and to 1/4th of an un
Re: (Score:2)
Unless they start putting phase array antenna's on APs to tight beam data to devices there is simply no avoiding time sharing channels... still for a given time on a channel it still gets more data through it, so even in congested situations it helps.
Re: (Score:2)
It helps against interference and attenuation ... but competing routers aren't treated as interference. Once they can beam it tight enough that the CSMA/CA algorithm is adapted to take it into account (ie. the spec allows one router to simply shout across a competing one with a directional signal) then things will change. I don't think beam forming with 4-8 antennas will be nearly tight enough to allow that.
Re: (Score:2)
Re: (Score:1)
802.11ac isn't out yet but I have little hope of it really helping. I live in an apartment building I can already see 50+ routers on 2.4 & 10+ ON 5GH.
I just don't see that much of a benefit unless the congestion avoidance is really better than 102.11n.
You are probably right. I think this invention will be of more help: http://hardware.slashdot.org/story/12/10/23/1946248/increasing-wireless-network-speed-by-1000-by-replacing-packets-with-algebra [slashdot.org] It works basically "repair data" in RAR archives for WiFi networks, so distorted/damaged packets can be recovered even if their checksums don't match.
Realism... (Score:3)
What's the intended range in realistic situations, 5cm?
Re:Realism... (Score:5, Informative)
Shorter range is a significant advantage today because it reduces interference.
Interference is probably the main reason for lack of speed and reliability in modern city apartment WiFi.
Re:Realism... (Score:4, Interesting)
And when you live in a house, like oh say 50%+ of the population, you will now have to purchase 3 wireless routers to ensure coverage, whereas previously only one decent one was enough.
Re: (Score:2)
Re: (Score:2)
Amateur radio licencees can use much higher power on some 802.11b/g frequencies (1500 watts, versus 1 watt), and so can use amplifiers that are illegal for the general public. However, it doesn't solve the problem at all - you need bidirectional communications for things to work. Boosting the AP transmit power doesn't make it receive any better. Adding a receiver preamplifier helps some, but since they boost the noise as well as the signal, they aren't as useful as you'd think.
If you need to boost a signal
Re: (Score:2)
There's a simpler solution - stop buying access points with inbuilt (ie, crappy) gain antennas.
Some have _PCB printed_ antennas that are just plain crap at everything.
Spend a little more money, buy something 2x2 or 3x3 with external antennas worth a damn, and you'll find your reach and throughput drastically increasing.
The number of houses I've been in lately with crappy wifi due to their APs having onboard antennas is just plain ridiculous.
Re: (Score:2)
Re: (Score:2)
RF propagation (and, thus, interference) is in no way dependent on the QAM modulation scheme and is the same for QPSK, QAM-16, QAM-64, QAM-256. Moreover, going to higher modulation constellations could force you to increase transmit power generating even more interference. Just my two cents from real world RF engineering
Re: (Score:2)
It'll be better than that, but not much. The killer problem is the average house would probably work pretty well WRT multipath distortion, but they don't need the bandwidth, and the office which needs the bandwidth is all steel framing and steel cubicles and steel beams and aluminum window frames which is going to multipath distort the signal into unusability. The TLDR is where its needed it won't work, and where it'll work its not needed. Whoops. Well back to selling kitty litter over the internet...
An
Re: (Score:2)
Re: (Score:2)
crypto (Score:3)
Re: (Score:3)
I should be able to setup an *open* access point with encryption these days so my users don't get their email passwords jacked. There is absolutely NO technical reason why this cannot be part of any modern specification. I will never ever use an open access point for this very reason...
Security: you don't understand it, so you're doing it wrong. Your problem is that you're too trusting. You don't trust sniffers, sure, but you trust the router you're connecting to, which is beyond daft. You should never trust some random jerkoff's AP. They could well be subjecting all of your packets to DPI to look for passwords after they are decrypted by the AP. Or you could use https or ipsec and then the data remains encrypted until it arrives at the intended destination, which is how you actually prot
Re: (Score:2)
So you're entire argument is that one should trust unencrypted public airspace more than (or as much as) one should trust a single router?
Less exposure is less exposure, and that's good. It ain't perfect, but nothing is.
Re: (Score:2)
So you're entire argument is that one should trust unencrypted public airspace more than (or as much as) one should trust a single router?
Mine entire argument is that one should not trust the encryption betwixt wireless ethernet card and access point to protect anything at all. Do not trust it to keep people off of your wireless network, and do not trust it to protect your email password, and do not trust it to protect your browsing habits. You may reasonably trust a combination of firewall and IPSEC provided you keep up with advisories and updates, and even that is plenty debatable. How do you define "trust"? You trust the tool to do what i
Re: (Score:2)
His argument is that you shouldn't trust an unknown router OR unencrypted wireless. Encrypted wireless just gives you a false sense of security.
The solution to the problems you mentioned is to phase out unencrypted passwords. In the meantime, use end to end encryption.
Re: (Score:2)
His argument is that you shouldn't trust an unknown router OR unencrypted wireless. Encrypted wireless just gives you a false sense of security.
No. Encrypted wireless gives you 100% protections against other wireless users. It is 100% effective at protecting from 99% of the likely threats (discounting the government and phone/cable companies). And I think that's worth something.
The solution to the problems you mentioned is to phase out unencrypted passwords. In the meantime, use end to end encryption.
Encrypted passwords don't come into it. Even with end to end encryption, encrypted passwords don't come into it. We're talking about data in motion, not data at rest.
Re: (Score:2)
Encrypted passwords don't come into it.
It's what we're talking about, you might try keeping up.
Even with end to end encryption, encrypted passwords don't come into it. We're talking about data in motion, not data at rest.
There is no difference. Data is data, bits are bits. They don't take on some special property because you send them through a wire.
Re: (Score:2)
There is no difference. Data is data, bits are bits. They don't take on some special property because you send them through a wire.
They kind of do:
http://www.schneier.com/blog/archives/2010/06/data_at_rest_vs.html [schneier.com]
We have been specifically talking about data in motion, where part or all of the journey is (or is not) encrypted. "Encrypted Passwords" implies the data at rest portion of the problem - for which the generally accepted best practice solution is bcrypt.
http://stackoverflow.com/questions/4494234/what-are-the-best-practices-to-encrypt-passwords-stored-in-mysql-using-php [stackoverflow.com]
Solving the rest portion obviously does not solve the motio
Re: (Score:2)
Bruce has been wrong before and he's wrong again. Cryptography solves one kind of problem: the kind of problem where someone might intercept your bits and glean meaning from them that you would prefer remain dark. That problem crops up both when sending bits across a wire and when bits are lying around unused. Believing that bits "on a wire" are different from bits on a disk is the same as believing that a business method "on the internet" deserves a different patent from the same business method without th
Re: (Score:2)
The needs and reasons for encryption are inherently different for data at rest vs. data in motion. Yeah, the end result is that you don't want someone to see them, but the vectors are so different it is useful to think of them differently. Data at rest need never be decrypted. When it is decrypted, it need only be decrypted once, and only in one location.
Data in motion is almost always unencrypted twice - once when it is being entered, and it *must be* again when it is used at its destination.
You often apply cryptography at a different time for data being transferred, but not only is that not always the case, but that's irrelevant when we're discussing interception of encrypted data. The only things that're relevant are how it's encrypted and whether it's encrypted when an attacker intercepts it.
Timing is i
Re: (Score:2)
Re: (Score:2)
Speed... Meh (Score:2)
Always obsessed with speed. Year after year all most of us want is better range and less interference, not more speed. More channels and frequencies are needed. Do that and it might have a chance at being interesting.
The law (Score:2)
Re: (Score:3)
I know it would require more bands... and that is precisely what is needed. What is taking so long? The FCC auctioned off billions of dollars worth of bands to companies for mobile phones/etc, and we citizens are still stuck with these few crappy little crumbs for one of our most important wireless technologies.
Re: (Score:2)
obsessed with speed. Year after year all most of us want is better range and less interference, not more speed.
Well no. I live in the boonies, and all I want from my wifi is more speed. I don't have interference to worry about, with no microwave and a 5.8 GHz phone next to a 2.4 GHz AP.
Re: (Score:2)
You are probably not "most of us", then :)
Consumer Fail / Backhaul Win (Score:1)
802.11ac will be a relative failure for the consumer market. While it will provide great speed benefits for very short ranges, the penetration og 5GHz through walls etc. will require lots of APs and won't be widely adopted.
However...
802.11ac for open space, line of sight, long range backhaul links sounds pretty promising.
ac is relatively useless (Score:2)
It's got more application in a server room than it does in the home. Range will be no better (and probably worse) than 5Ghz on your 802.11n routers and the amount of 5Ghz frequency it requires is simply put: ALL OF IT. So you won't be able to place multiple routers in an area, because they'll stomp on each other either at the AP or the endpoints.
This is a standard for (at best) a home audio/video system where all the components are nearby and for a server rack where you use wireless as a second network
Re: (Score:1)
I manage an installation of nearly 5,000 APs. We have about 800 in the larger buildings. Yes, we do get basic 802.11n speeds on 5GHz, but 2.4GHz is a crap shoot. Way too much interference both non-802.11 and co-channel from neighbor APs.
We still have trouble with 5GHz as we are seriously out of channels with 20MHz bands. We can't bond them to 40MHz. So, 802.11ac is functionally useless. I have told that to vendors marketing 802.11ac products to us. Without more channels, we can't bond.
At home, I can probabl
Re: (Score:2)
I have no idea why you were downvoted. What you've said is entirely accurate.
What's in it for me? (Score:2)
All my WiFi gear still uses the 802.11b range as I get better speeds from it then the .11g mode and don't even try N mode as I don't have any adapters that offer/use it. Simply aint worth the money due to to many routers on the same damn frequency. Hell I'm even looking to switch back to the old 900Mhz band for cordless phones due to everyone having moved to the 5Ghz models. Better range and no interference from others. All of this crap in the same band simply chokes 802.11n speeds to less then what I see
Re: (Score:1)
Yeah? Well, I'm still using 100BaseT for everything. I'd tell you to get off my lawn, but I'm pretty sure there's a mean old geezer dozing off around here somewhere who still uses 10Base5 or maybe even token ring, and if I yell at you too loudly he'll probably wake up and kick us both off the grass.
What about range? (Score:2)
Who cares about speed. 802.11ac will give you better range. 4x4 will give you better range because of beamforming.
Re: (Score:2)
Uh.................. only if they support 4x4. And that range will be very limited. 5.x does not penetrate obstacles well.
Show me a single 802.11n router today that fully support beamforming. Even the Ruckus and Wavion gear are only using part of the 802.11n spatial multiplexing capabilities.
What the hell... (Score:1)
Re: (Score:2)
...happened to the rest of the alphabet sequence? No M?
You jest! BUT...
The IEEE suddenly realized they'd have to do like in apartment buildings: Normally named from 1A through Z their units beyond the first 26 combinations require a wrap-and-restart at 1AA for the rest of the units on that floor.
"That's MADNESS! I've never seen these other 20-something protocols!!" a curious slashdotter might say... Yes, there's really a lot of networking protocols AND they're just not ALL in the domain of Wifi tech. It's networking in general. The power of "QoS" we consider
Re: (Score:3)
Re: (Score:2)
Already doing that. My Freebox in in one corner of the apartment & I have an old WRT with the same SID/password in the opposite corner where multiple walls & neighboring wifi APs render the freebox wifi unusable. I can thus pass off from the freebox & the WRT transparently but that doesn't change the high noise floor & interference from all the other nets.
Re: (Score:3)
None. Most of them are about reducing congestion problems, typically through reducing range (like 2.4GHz to 5GHz shift) or various antennae and beamforming solutions.
Your problem is the exact opposite, and typically handled by installing additional WiFi routers or amplifiers. Your problem has been long solved, but it would appear that your motel is simply not interested in investing in solutions.
Re: (Score:1)
I already use 2 APs & own the apartment I live in. Don't assume that because you live in a motel that everyone else does.
Re: (Score:2)
Get some extra wlan routers, put them at intermediate points and bridge to different channels to extend the range (repeaters make inefficient use of channel bandwidth). Not entirely trivial to setup but perfectly doable even with cheap routers.
Re: (Score:2)
I already use a WRT that has the same SID/password as the 11N AP to cover some parts of my apartment. That doesn't change the high noise floor/interference from all the neighbouring APs.