There is 802.11ay expected in 2019 as well, a step up from 802.11ad using the same 60Ghz. Up to 20Gbps, 3x the rate of 802.11ad.
I really really want 802.11 ay and ax to come faster, it will truely bring us to the world of Wireless with the speed we expect from 1Gbps. Wireless Display to TV no longer be slow of choppy with quality degradation.
Someone is going to make a killing on tiny, inexpensive RJ45 60GHz access points and switches that support PoE. Killer use case: stick it into a wall, and everything in the room gets gigabit speeds. No wall warts to deal with.
5ghz AC doesn't go through my plaster walls now. I've got two walls between me and my Airport Extreme and about 10 physical feet and have half signal showing on my phone.
In this case, what they mean is that the signal will be truly line of light. As you go up in frequency, radio starts to behave more like light so while you may get some convenient reflections here and there, for the vast majority of the time, if the antenna doesn't have a clear path to the transceiver, you won't know there even is an access point.
I suspect this can be quite convenient for various places where you want to limit access to individual rooms etc, but don't want to deal with the tangle of cables and ports that a wired network would require.
If you want a good 5ghz AC setup buy a couple of good repeaters with seamless handoff and put one in each room that's what 5ghz was mainly designed for.
Airport is a very crappy 5ghz ap to begin with so is the Google elcheapo-de-slink.
As I always request in these threads, I'd like someone to point me to a peer-reviewed paper that shows 5GHz being attenuated more by building materials than 2.4GHz.
The usual culprit is that people put their 5GHz networks on the old "low power channels", which nearly every router I've ever seen chooses by default. Half the power, half the signal in the next room, of course.
I see this is 900MHz, how will that fair in countries that utilise GSM900 as that may certainly be a restriction in adoption for some countries due to frequency clash, would it not?
GSM900 is 890-915 MHz; 802.11ah gets around this by using region specific frequencies that sidestep GSM900:
902-928 MHz in the US (26 MHz bandwidth)
863-868 MHz in Europe (5 MHz bandwidth)
916.5-927.5 MHz in Japan (11 MHz bandwidth)
Limited bandwidth, especially in Europe, will to lead to both higher power consumption and lower real-world range.
>The PHY layer will allow devices and APs to operate over different sub-1GHz ISM bands, depending on the country regulations: 863-868 MHz in Europe, 902-928 MHz in the US, and 916.5-927.5 MHz in Japan. China, South Korea, and Singapore also have specific channels.
It also means they'll be less channels available, so more noise. These frequencies are commonly used by things like door bells, burglar alarms, garage door openers, car locks, etc. I'm more worried what the effect will be on them.
My main issue here is that each new wi-fi standard requires new hardware. I remember reading something within the past few years about revolutionary new wi-fi antennas that will be able to adapt to any frequency such that new hardware wouldn't be an issue. I wonder if anyone has seen anything similar?
Each new hard drive standard has required a faster motherboard to take advantage of it.
WiFi standards are backwards compatible in the sense that they work with older hardware to send at the maximum speed that those radios were capable of transmitting.
That's simply not true. Each standard doesn't require a faster motherboard, only some changes to code and possibly a different communication frequency. Changes to code do not require different hardware while changes to the communication frequency do.
The crux of my initial comment is that it would be awesome if we could some how get around the last issue and not have to replace hardware in order to use a different frequency, which @walshemj points out is a physical impossibility, but I believe I read an article a while ago in which someone had gotten around this?
The antenna is part of it, but it is also the radio. A lot of the hard work of modulating the wifi signal is done in this hardware, and low-level firmware on the device.
We do have Software Defined Radios, which practically allow you to tune to a much wider range of frequencies, and speak nearly any wireless protocol though.
That's not going to happen antenna design is restricted by the physical laws governing radio transmission - you wont be able to make a magic antenna that works across all wave lengths.
I think it would have to do far more with the green board and not the antennas. Especially at the home owner router level your not going to put the effort into the back plain of the router. And while most of them aren't really general purpose computers in that, a good part of the functionality is going to be baked into the various chip hardwares on the board. One might be able upgrade the OS but changing the functionality of the radios probably totals the box.
Plus, what would they sell you next year? To be sustainable, the company would have to be mostly a software company that people are willing to pay license or upgrade fees for increased functionality or performance. There may be a niche for that, but most people won't upgrade their router or CPE for free. At scale, I doubt that would pan out, especially with increased, upfront hardware cost.
Using an admittedly expensive Asus RT-AC88U with my MBP connected over 5ghz, I can get ~900mbps on speedtest.net using my gigabit connection.
Example test below [0], admittedly the downstream results aren't the best but I've got some rather large downloads and streaming going on right now. Upstream shows the capability though.
That is pretty damn insane for real life speeds. Most reports I've seen say ~400.
I must admit I'd probably not upgrade anyway though. 100mbps is enough to carry a couple 4K video streams in parallel. So not going to outgrow that speed anytime soon.
LAN also outperforms wireless of far higher speeds for certain work, in my experience. 10mbps LAN readily trounced 200mbps wireless for SMB file sharing, for me.
I attribute it to less dropped packets, less interference, and full duplex, despite lower bandwidth.
Good wifi deployment shouldn't lose packets ever. 802.11g (54MBit) easily reaches 20Mbit so your deployment had to be utterly terrible for 10Mbit Ethernet to outperform it.
Don't forget that Wifi needs to be properly positioned to work quickly.
> Good wifi deployment shouldn't lose packets ever.
You lose packets to radio interference (or collisions). And SMB in particular reacts poorly to packet loss.
You're also assuming that "good" wifi deployment is a reasonable expectation. People just buy that blue Linksys thing and put it near the jack for the modem. If you're going to run wire all over the place for access points then you might as well run wire all over the place for ethernet.
>Don't forget that Wifi needs to be properly positioned to work quickly.
And if my properly positioned you mean "In a faraday cage with only the AP and your station", you'd be correct. The issue is real world WiFi doesn't ever work that way. There are very few places where you don't pick up at least one or two neighbors using at least some of the channel bandwidth you need. Even inside your own wireless domain you have the hidden node problem, forcing wireless domains to be broken into small cells using low power greatly increasing the costs of deployment. Add that a huge number of cheap APs and computer drivers flat out suck and the average non-professionally installed wireless user is not going to see 50% of the bandwidth they should.
Thing is, you really need the "300Mbit" rated 802.11n to get reallife 100Mbit in good conditions. In a lot of cases only 802.11ac capable devices really go over 100Mbit - the 1600Mbit stuff is mostly just marketing speed unreachable in practice.
60GHz (which limits it to basically one room, as it's quite readily attenuated by even air).
Living in a condo tower where I can see -- and have my network interfered with by -- 70 other wifi networks, I consider this to be a feature rather than a disadvantage.
This kind of problem should be solved by transmission power control. Power is capped by standard to avoid reaching nash equilibrium. Also, Wi-Fi has various OBSS coexistence mechanisms to detect overlapping networks and cooperate with them to reduce power simultaneously for mutual advantage.
Have any security researchers investigated the possibility of abusing these coexistence mechanisms for malicious purposes? Could be worth checking out, I think.
On many networks you can just forge 802.11 disassociation frames and drop people off the wifi altogether. (In fact, enterprise wifi vendors will sell this to you as a feature - the premise being that you'll use it to keep corporate-owned laptops off unapproved access points - but there's potential for FCC trouble even there.)
Here's an interesting attack: Suppose there are three non-overlapping channels and 70 APs, most/all of which try to pick the least congested channel. Can you persuade 67 of the 70 APs that a particular channel is very congested and best avoided, while leaving that channel uncongested and usable for the few remaining APs? Can you do so without client cooperation from the three favoured APs?
> That said, at 60GHz of course bandwidth is awesome.
The carrier frequency doesn't tell you much about the information capacity of the channel. The width of the frequency band used is key, which is why bandwidth is often (incorrectly) used as a synonym for data rate.
60GHz is the central frequency, and tells you nothing of the width of the frequency band. The ratio of the width of the frequency band to your data rate (amongst other factors, including antenna gain and processing gain) tells you your signal-to-noise ratio.
I have a Intel WiGig laptop + dock that uses wireless ad under the hood.
The bandwidth can be used for much more than just network transfers. I've personally had 2 1080P displays playing video + USB3 external hard drive transferring at max speed + Gigabit Ethernet transfers using the dock without anything skipping a beat.
Wireless ad docking + wireless charging on a laptop will be absolutely amazing.
Bullshit, your device is at fault and not the protocol. I bought a 802.11ac dongle with external 10dbi gain antennas and have been enjoying extreme speeds and connection strenght throughout the house, not to mention to the routers of 30 of my neighbours.. And that's only 5GHz. For comparison, at my desk three walls away I used to get about 40-50nW signal level off my built-in adapter and antenna in my Lenovo laptop. With the new setup - 60mW... The router sends at 100mW. We're talking pushing 90MB/s+ constantly no matter where in the house with absolutely no connection drops.
If you buy a cheap AC router and expect a tiny built antenna to max out the bandwidth you will be disappointed.
I get over 1Gbit real tested bandwidth between my main desktop using a decent dongle with external antennas and a high end router.
Even iscsi works well with this setup the NAS is connected on 10gig Ethernet to my router.
With MU/MIMO you can also get full duplex if you set it up.
The dongle is built into the motherboard and it's connected via pcie.
Before that I had an asus pcie x4 wireless card which worked just as well(maybe even better).
> ... while also not being particularly constructive.
Eh?
The guy mentioned that he switched from using the built-in antenna on his laptop to a honking external antenna and now gets 720mbit/s performance anywhere in his house. That's quite constructive. I guess the only way it could get more constructive is if he mentioned the model of hardware involved... but I expect that you'd get similar results with any AP capable of 720mbit/s transfer to a single client. If your AP is fast enough, then it's all about SNR.
>but I expect that you'd get similar results with any AP capable of 720mbit/s transfer to a single client.
One should expect that, but almost no one uses wireless that way. If you have just a single client the most efficient method is to run a cat5 cable and get 1000mbit/s at a much lower cost than putting in expensive AC equipment, barring an expensive engineering issue of making a cable reach that place. The problem with wireless is everybody has to play nice and be able to see each other or it quickly falls to the 20% actual usable bandwidth rather quickly. Do you have neighbors with a wide bandwidth AC router and healthy streaming video habit? Don't expect to get max data transfer rates. Do you have lots of portable devices, especially cell phones that like to go into low power mode, or get set in places where they can reach the router but become hidden nodes to other clients in the network?
Speeds will vary greatly depending on whether proprietary booster tech is in use, which in turn will depend on the combination of hardware both on the sending and receiving side.
oh and while we're on the topic of bullshit (seeing how you were so blunt to go there) - signal levels are measured in decibels, not nW. nW is a measure of power fed to the antenna. Your comment is about as plausible as someone claiming their car is 40 gallons fast.
RSSI -- Receive Signal Strength Indicator -- is reported in dBm, which is a log-scale unit of watts. +20dBm is approx. the max transmit power for WiFi, and equates to 100mW. 1nw is eqivalent to -60dBm which is a pretty decent signal strength for WiFi, but won't provide max data rate. By comparison, the minimum power to operate at 1Mbps is about -98dBm.
Then again, receiving at 60mw (+17dBm) would overdrive any of the WiFi devices I ever tested and path loss of just 3dB (assuming +20dBm transmit power) is not achievable over the air, even with a near-field antenna. Keep in mind many devices actively tamper with the RSSI value reported up through NDIS, so while it's possible for a card to report it is receiving at +17dBm, unless you have hard-wired the hirose connectors between AP and STA it's not actually the case. Nor would a receiver work very well with so little path loss in any case, it would be completely over-driven.
On the flipside, is it ok that we are still shipping 1gb lan ports like 20 years after they became commonplace? I'd hope that at least enthusiast class hardware would start using 10gb at some point.
I must admit this is the first I hear of ad standard so I'm a little concerned that by the time I've got the tech in hand my internet options are already at the next level. (I should add that its all relative - 100mbps is just fine for my needs to be honest & will be for a long time)
Its a touch ironic. In my previous country it was the opposite way round - nobody could imagine a situation where the internet link could saturate the wifi. No I've got the opposite problem...
You'll get about 10 metres in an office with cubicles with chest high partitions. Page 6 of the paper "A High Speed Wireless LAN" by Skellern, Lee and others in the Jan/Feb 1997 edition of IEEE Micro has a diagram showing coverage for OFDM in a typical office at a centre frequency of 40GHz. 60GHz will be worse than this, as oxygen in the air absorbs 60GHz quite stongly [2]. I'm not sure how the power transmitted by 802.11ad compares to the above measurements, but the measuremensts in the paper were taken with 60GHz WiFi in mind, so they should be representative.
The laser analogy is definitely accurate given today's technology. I just don't think they'll be able to productize something that has that high of an antenna gain. I'm really hoping and thinking that we'll see some smart antenna advances.
It will depend on the material, but I don't think 802.11ad can even penetrate through sheetrock. Honestly it may even be so bad that a sneeze could knock your connection out for a few seconds.
"Twice the range".. Is that a doubling of the radius or a doubling of the diameter (edit: oops, I meant area) of a ~ circle surrounding the access point?
Am I being spoiled thinking either of those options aren't super impressive for a whole new standard? Where's the 10x (order of magnitude) improvement? Will it take 3-5 further complete iterations of incompatible standards? :)
As usual with Wi-Fi you won't get close to theoretical range and distance in practice. 802.11ah includes support for relay access points [1]. Most probably it is a requirement for "nearly doubled range".
> Am I being spoiled thinking either of those options aren't super impressive for a whole new standard? Where's the 10x (order of magnitude) improvement?
802.11ah is not the "mainline" standard. 10x improvement is left for 802.11ax. 802.11ah is for IoT applications: low bandwidth, low power consumption, high distance.
I don't believe we really use cordless phones in the 900MHz anymore and I believe the consortiums determine which spectrums are sufficiently unused before targeting them.
Verizon and T-Mobile have some LTE spectrum in the 700MHz range for building penetration and range so I expect to see great things at the 900MHz range for lower power devices.
It's not just cordless phones though. Our baby monitor says it uses 900MHz. I don't know how many types of devices use 900MHz but it seems crowded to me.
I recently bought an 802.11ac wifi adapter promising a nice theoretical 600Mbs in the 5Ghz band and all I was able to get from it was a practical 65-70Mbs (at 2 meters from an 802.11ac router) using the best drivers under Windows. Under Linux the (5Ghz)AC mode didn't work (buggy driver; no vendor support) so I was back in 802.11n mode. Quite underwhelming after spending 200 euros on router and adapter. I will definitely not "early adopt" wifi 802.11ah whenever it appears..
If you get a reasonably good combination of hardware/drivers, 802.11ac works well. I'm able to get 407.99 Mb/s Down, 462.39 Mb/s Up on a 2012 Macbook Air + RT-AC87U WiFi Hotspot on 5 Ghz to speedtest.net in Singapore.
I can confirm similar local measurements with MacBook Pros and a D-Link DAP-2950 across a room with a wood partition blocking line of sight. If the AP is advertised as 600Mbit/s, chances are it only supports 1 or 2 streams (MIMO antennas) per band. Most laptops only have 2, but even so the signal quality will be much better wih a 3- or 4- way MIMO setup on the AP end. (The D-Link has 6 antennas, so 3-way MIMO for both 5GHz and 2.4.) In addition to number of antennas, orientation also makes a big difference.
Finally, wireless access points seem to be one of those things where quality varies a lot. I hated wifi for its unreliability until I bought a decent (professional/small-to-medium-business grade) AP 8 months ago. Not a single connection problem since, even with the notoriously affected Mac hardware and OSX software versions.
I get ~180Mbit with my Asus RT-N66U (802.11n) with 2x2 clients and the RT-N66U is capable of 3x3. A friend recorded over 700Mbit with his Nexus 9 and an Asus 802.11ac AP.
You should be getting muuuch faster speeds. snbforums is a good place for wifi help.
My router is an asus rt-ac68u and I have tomato installed on it (I wonder if that's the problem..). The wifi adapter is netgear a6100. The testing was performed in an internal network. I will keep troubleshooting.
Stock firmware is known to have the best wireless. If you want a more featureful alternative firmware, try AsusWRT-merlin. It is practically an enhanced stock.
I tried merlin before and I'm really torn.. tomato does some things clearly better (dns-crypt, ftps, file sharing, bit torrent, UI, customization) and I wouldn't give up tomato if it wasn't strictly necessary.
IPv4 lets the MTU get as low as 68 (but you're only sending 8 bytes of data per packet); IPv6 prevents it from getting any lower than 1280, but permits "link-specific fragmentation and reassembly" as long as IPv6 itself doesn't see any smaller fragments.
IPv6 also permits (but does not require) jumbograms that make full use of the 32-bit packet size field, but across a wireless link, having to retransmit 4 GB at a time seems like a bad plan... this only seems useful for short-distance wired connections among hosts that know they've got a quality cable between them.
The 2.4 GHz and 5.8 GHz audio protocols seem to not be wifi-based. (Which makes sense, you don't want them to be IP-based.) They just transmit uncompressed analog audio over probably something like FM.
900 MHz headphones seem to be readily available, and to have been readily available for years.
That's utterly incorrect all wireless headsets these days use digital audio streaming of some sort doesn't matter if it's proprietary like PurePath Wireless (one of the most popular ones because it's TI tech which means allot of very cheap and vailable SOC's), AirPlay or more open like BT A2DP.
No they don't use simple radio, all of them will use a packet based protocol either proprietary or some sort of a standard digital (audio) streaming protocol.
Encoding and packets don't add up any substantial latency just look Bluetooth A2DP headsets and AirPlay speakers (which uses Wifi as a carrier) headsets.
If your headsets did use just normal 2.4ghz radio like say a basic walkie talkie the downside is that you would've been able to hear any other transmission on that band e.g. an old baby monitor(new ones are digital too).
Hopefully nowhere and never. These are bands unsuitable for high bandwidth applications and we certainly don't need more of the public spectrum polluted by things like cheap chinese RF headphones.
802.11ah: Better range, lower bandwidth
802.11ad: Better bandwidth. Doesn't go through walls.
802.11ax: Successor to 11ac expected in 2019