How Wi-Fi Works | Hacker News Reader

How Wi-Fi Works

verizoninternet.com

308 pointssharjeelsayed92 comments

This is (good) SEO linkbait. Someone at Verizon got $10k to spend getting it created by saying it will boost organic search traffic to Verizon. Right now (during link-building phase) they keep the page completely separate from rest of site. Later, (after most links are created) they'll change it. Not sure whether the goal is just to generally build authority to VerizonInternet, or to get this URL ranking for wifi keywords. (Seems more likely the former.)

That page probably cost a lot more than 10K.

Absolutely. They did a great job too

Browser could not initialize WebGL For more info visit blend4web.com/doc/en/problems_and_solutions.html#problems-upon-startup

Maybe they could have done a good job instead of a great job and this page would actually work.

Not mention the font size issues and overlapping text, luckily they provide images at the bottom that do not have the problem.

When you have to provide images at the bottom as a workaround, it usually means you know there are issues with the page and the job is not that great.

Then again, as pointed in other comments here some of the technical information is incorrect or just plain wrong.

Would it be accurate to assume this is gonna be used as ammo in the net neutrality debate of the next few weeks?

Are you saying they'll change the content later, and won't lose seojuice for that?

The promoted embed snippet actually points to a different URL: http://www.verizoninternet.com/bookmark/guides/how-wifi-work... which contains a bunch of site gunk for their "bookmark" blog, where they post a lot of content (and probably some net "neutrality" articles coming soon?)

As hatsunearu said, the radio modulation described is grossly incorrect. WiFi never uses 8-PSK (encoding 3 bits per symbol). 802.11n and 11ac encode 1/2/4/6/8 bits using a BPSK/QPSK/16-QAM/64-QAM/256-QAM symbol (256-QAM is for 11ac only). The modulation scheme is negotiated based on signal quality. Here is a quick reference: http://mcsindex.com/ (MCS = modulation coding scheme) On Linux you can find the MCS negotiated with "iw dev wlan0 link | grep -i mcs"

14 channels are defined in the 2.4GHz band. For example channel 6 is centered on 2437 MHz. Each channel is 20MHz wide and divided in 52 "data" subcarriers, each occupying a different frequency and spaced out by 312.5 kHz (52 × 312.5 kHz is less than 20 MHz because there are "control" subcarriers and additional spacing.) So 52 different symbols can be sent in parallel at the same time, which is what we call OFDM https://en.wikipedia.org/wiki/Orthogonal_frequency-division_... (basically, I'm simplifying!)

Remember this is for just 1 channel. So with 14 channels each composed of 52 subcarriers, we could have 728 symbols transmitted at the same time. If they are 256-QAM symbols that's basically 728 × 8 = 5824 bits being transmitted at the same time in the air. And they will all be received and demodulated independently. This high level of parallelism of OFDM is how WiFi can achieve very high throughput.

Then, with wide channels of 40 MHz, which basically aggregate two 20 MHz channels, we get a few more data subcarriers because we don't need as many control subcarriers so a few of them become used as data subcarriers. Hence a 40 MHz channel will have not 52 × 2 = 104 but actually 108 data subcarriers. And 802.11ac defines 80 MHz and 160 MHz channels with respectively 234 and 468 data subcarriers.

Let's calculate the maximum usable throughput of a single 802.11ac 160 MHz channel using 256-QAM modulation... It sends 468 symbols at the same time on 468 data subcarriers. Each symbol encodes 8 bits and takes in the best case 3.6us to be transmitted: 3.2us for the actual symbol + a short guard interval of 0.4us (the GI is normally 0.8us but can be a short GI of 0.4us if negotiated). The raw physical bitrate is:

1/3.6e-6 × 468 × 8 = 1.04 Gbit/s

However there is a mandatory error correction which is 5/6 in the best case so the actual usable bandwidth is:

1.04 × 5/6 = 866.67 Mbit/s

> 1.04 × 5/6 = 866.67 Mbit/s

This is bang on what most hardware is able to achieve.

I am curious Apple haven't jumped into 512-QAM yet. Partly stopped me from upgrading my fiber to 1Gb/s.

Nobody uses 512-QAM for WiFi, and for some good reasons:

1) it's not part of any current or draft standard;

2) it only encodes 9 bits vs. 8 bits for 256-QAM. Plugging this into GP's formula gives us:

1/3.6e-6 × 468 × 9 × 5/6 = 975 Mbit/s

3) current WiFi QAM constellations are all square [1,2], and there is no way to arrange 512 points into a square.

4) that would require an unrealistically clean channel, and give no advantage over lower-order modulations otherwise [3]

[1] https://dsp.stackexchange.com/questions/31607/why-are-qam-co...

[2] http://www.radio-electronics.com/info/rf-technology-design/q...

[3] https://upload.wikimedia.org/wikipedia/commons/thumb/6/68/QA...

Ok. I think I’ve meant 1024-QAM.

Note that, although the 2.4GHz spectrum is formally divided into 12-14 channels (depending on local regulations), these are very narrow channels; in practice there are only 3 non-overlapping 20MHz-wide channels. This is a small fraction of the width of the 5GHz band.

There are 3 non-overlapping 2.4Ghz channels (1,6,11) for 802.11b, because of the channel spectrum shape. Not only "b" uses 22Mhz channels with is just a bit too wide, but also, due to the way how single-carrier PSK modulation works, especially on older hardware, "b" has considerable amount of spurious emissions adjacent to the main carrier that widen it even more - could be seen here as smaller "hills" to the left and right [1],[2].

802.1g/n/ac can easily have 4 non-overlapping channels (1,5,9,13) because (thanks to OFDM) channel spectrum is much neater with rather square 20 (40) Mhz channels with practically no energy outside [3].

Yet, everybody's stuck with 1,6,11 channel scheme which is wasting precious bandwidth. [4] (middle graph) Notice gaps between the channels that could be eliminated by moving 6->5 and 11->9, and gap on the right where channel 13 can fit after that.

[1] https://villagetelco.org/2009/11/rf-hacking/

[2] http://ecee.colorado.edu/~ecen4242/wlanb/index_files/image02...

[3] http://rfmw.em.keysight.com/rfcomms/n4010a/n4010aWLAN/online...

[4] https://www.codify.com/wp-content/uploads/2009/07/Central-Ro...

> Yet, everybody's stuck with 1,6,11

Reminds me of:

https://blog.xkcd.com/2009/09/02/urinal-protocol-vulnerabili...

"How Wi-Fi Works" --> 503 Service Unavailable

Seems about right.

Hahaha, I was about to post just that. Made me laugh. Reminds me of my time at university... during breaks and boring lectures, WiFi was impossible to use because of the load. You could probably measure how boring a lecture is by recording WiFi latency and stability parameters.

I've got that once it requested WebGL access and I've clicked "Never for this site". Refreshed and worked fine.

Verizon is trying to give you the authentic experience of how your WiFi works when you're on their network.

Getting a 404 now. Any idea where the page went? (Here's an archive link: https://web.archive.org/web/20170723145452/http://www.verizo...)

Are you sure it was a 404? I'm getting a 503 (Service Temporarily Unavailable). Not that it really matters...

This is beautiful, but probably still too complex for most of their customers. I wonder what their motive in putting this together was, as it must have been very expensive.

Maybe the fact that as an internet provider it's their duty to explain how what they're selling works? Granted, "internet" doesn't mean Wi-Fi however their equipment includes a wireless access point so it's fair for them to provide documentation about how that works.

> it's their duty

Not at all. No more than a car manufacturer has a 'duty' to explain how a car or a car engine works.

>No more than a car manufacturer has a 'duty' to explain how a car or a car engine works.

It is their duty (car mfgs) in the sense that they need the customer to attribute value to their newest technology offerings.

Honda made damn sure that consumers knew what 'VTEC' was when it was new. Toyota made sure to throw 'Hybrid Synergy Drive' around all day; along with an on-dash animation of the flow of electrons. We're all hearing all about how Teslas' auto-pilot works.

I don't know if it's a duty unless you consider it as a duty to their shareholders to generate value or profit wherever possible.

VTEC and Synergy Drive are marketing terms, not established aspects of modern technology that consumers already know they want to purchase.

> they need the customer to attribute value

Yeah, and it's my duty to eat?

Not legal duty but moral duty.

And car manufacturers do explain - via handbooks and manuals and other instruction.

Yeah, it is beautiful.

Unlike their overcomplicated phone plans and billing statements which are deliberately obscure and their customer service which they run like a 2-bit boiler room operation.

Specially obscure. Which is like scamming, because most people won't try to understand it.

it was nearly too complex for my pc! I had hardware acceleration disabled in Chrome and it completely pegged my CPU out. the animations barely worked until I re-enabled it, but then the page ran out of memory. I ended up having to look at it on my phone.

I'm not convinced the manipulable 3D modem rendered in-browser in realtime really added a lot of value for me, but it's pretty cool I guess.

Just a reminder that Verizon its the biggest lobbyist against Net Neutrality[0] and if you do support it then it's probably wise to stay away from their services as far as possible.

[0] https://www.dailydot.com/layer8/lobbyists-net-neutrality-fcc...

Mirror: http://webcache.googleusercontent.com/search?q=cache:http://...

google cache doesn't do the page justice. most of the value is in its visualizations.

True, but better than the 503 I was receiving.

The cache now has the images, though it didn't when first posted. Honestly, the Google cache is _better_ than the original as it doesn't have the CPU-killing animations.

No images for me :/

The Archive of it is better https://web.archive.org/web/20170723145452/http://www.verizo...

These guys should do a followup "How Load Balancing Works".

Speaking of how wifi works, I learned something interesting about wifi and Verizon's partner in many things, Comcast: Last night I notified my home Internet acting funny, and learned that the admin interface for my Comcast router had username "admin", password "password". SMH.

If you want to hear something else scary, I dumped the firmware on my modem a little while back and started exploring it:

https://twitter.com/joshumax/status/860712276717748225

Did you need to do anything special to get shell access?

Just the basics.

Plugged in my 'scope and started probing some debug headers that looked a lot like they'd be for UART, check if one is Tx and is sending out data, figure out the baud rate, hook up Rx, Tx, and GND on my UART dongle to the correct headers, and modify the bootsting in Cisco preboot to spawn a serial console which landed me into busybox as root :)

> Just the basics.

Followed by the use of >=$100 of hardware and some not-beginner skills. Snark aside, I highly recommend anyone remotely interested in what is going on in your modem/router to have a go at this. You don't need the scope if you're okay with trial and error and it's pretty hard to break anything as long as you don't connect the 3.3/5V line to start with.

I think he means people-who-work-with-routers basics :)

Nice, I would have assumed there would be additional security for production units honestly.

... which is "okay", since you can only access it (the admin web console) from within your wlan/lan (and not the internet) and ofc you can/should change it during setup

Most people (esp non-HNers) don't, they just let Comcast set it up. To make matters worse, they set the SSID to my last name, and the password was my address. Maybe that's one-off, but if standard, seems problematic.

I haven't seen a telecom-provided wireless setup in forever that didn't have the default password be fairly long and random, printed on a sticker on the back of the router.

Seems fairly secure to me.

When I had AT&T it was definitely a more secure setup.

I can confirm that two different ISPs have done this with my initial WiFi setup over the last few years. AT&T made it the initials of everyone staying in that house with the password set to their 800 number for service calls. Time Warner made it one person's first name and the password was his cell number.

On the other hand, a Midcontinent Communications (aka Midco) tech told me the password I wanted to use wasn't secure enough and brainstormed with me for a couple minutes on good SSIDs and passwords while he showed me the web admin interface on my laptop. I was very pleased with his visit and called the local office afterwards to pass along kudos!

> ... which is "okay"

Nope. Anyone within range of your wifi router can connect to it and most possibly the first username and password that they will try is admin\admin or admin\password

The parent posts here are talking about the admin interface to the router, not the wireless password. While it's technically true that anyone within range can connect, they can't authenticate without the wireless password, and so cannot access the admin interface.

I mean, I just got a 10-gig router,[1] and the stock username/password was "ubnt"/"ubnt." It's always the installer's job to set up a new username/password.

[1] https://community.ubnt.com/t5/EdgeMAX-Stories/EdgeRouter-Inf...

Why not generate a long password and print it on the backside? My DSL provider can do it (for my DSL router)

People might be less inclined to change it. And they should change it, because some firm knowing your password isn't safe either.

It's worth noting that all the ISPs that encourage you to change your password have a separate maintenance account ("backdoor") with its own password.

All the WiFi routers I've bought in recent memory do that for the network connection password, but not for the admin page.

My NEC router had no default password and required you to make one up yourself on initial setup.

ISP's like spectrum discourage dispatching techs and opt for customer setup. And every model they give you has an 'admin/admin' or 'admin/password' setup.

I think the difference here is that someone buying Ubiquiti is knowingly getting a slightly more in depth initial setup.

Your grandma is getting a Comcast router and probably doesn't know it can be logged into and configured or how to do it.

ER-XG is quite far from home WiFi router though :)

UBNT's new-ish home router series (AmpliFi) doesn't have default username/password- it needs to be set-up before using. I do think it's possible to have an open-network default configuration, but the LCD will nag you to set-up the device, and the first step of the setup is choosing a password (both for management and WiFi).

Disclaimer: I work at UBNT.

Also there was a time when Verizon's modem/router had generated passwords based on the SSID https://aruljohn.com/fios/

Why can't it be "admin", "<random string of x number of nums chars etc>". Could even print it in the bottom of the device.

It's bad practice to assume things are set up properly by someone else. This applies to more than just computers, fwiw.

My ISP just set me up with gigE fiber. The installer was a bit clueless, didn't really want to provide me with the gateway IP for the router. Then I looked at the wifi settings; a 2.4Ghz and 5 Ghz setup, both with SSIDs that included the provider's name as a prefix. The password was a 9 digit password, all numerical...

I called the support line and got through quickly to an admin who could change the password and SSID for me (unless you pay for a public IP, the mgmt interface is locked down). I mentioned that having such a short, all numerical password would mean that any access point they set up would be trivial to crack. Just wardrive looking for similarly named access points, and you'd be able to jump on their connection in just a few minutes. He didn't seem to care, which is too bad.

Does anyone know, if I'm on a WPA2-PSK wi-fi, do other devices that are also on the same network can "sniff" my traffic. For unprotected networks it's obvious, but what about protected?

It wouldn't be a good system if they could. You have to worry more about wired Ethernet which can also be encrypted but rarely is.

As far as I know, they can [1]. Once you're connected to the network you can sniff out everything from clients that connect after you.

The standard response to this is that you're safe since the eavesdropper needs to know the shared key. You can look into setting up WPA2-Enterprise if you're worried about that. FreeRADIUS doesn't seem particularly hard to configure.

[1] https://superuser.com/a/156969

Yes, other devices on the same network (using WPA2-PSK) can sniff your traffic. It's possible to prevent this with WPA2-Enterprise.

Yes, if you know the PSK (=password) and can capture the initial handshake (which is easy, since you can just force-disconnect a client so it has to do a new one) you can decrypt it. (If I remember correctly, Wireshark has this built in, so you can try it for yourself if you are curious)

I'm assuming you mean a malicious device can force a deauth on another client (or more usually, all other clients) and then capture the packets as they reconnect. If so, is there a way to detect this? Is there any way to protect against this? I'm assuming client isolation makes it more difficult.

Well, the easiest way to protect against it is making sure an attacker doesn't know your wifi password.

Of course, but I'm working in a big office with office-private wi-fi with one password. And who knows what viruses my colleagues have.

Also, I have a pretty special audience of hackers here, so I wouldn't be surprised if someone actually tries it.

Same goes for various events and conferences.

Apparently since I last looked into it, "Protected Management Frames" from 802.11w are a bit better supported (in non-professional APs), which solve this issue by not allowing "anonymous" deauthentication. (requires support on both clients and AP though)

You of course can monitor for deauthentication packets, but unless you know when/if your AP is sending them during normal operation you can't make sure that an individual occurrence is an attack or not. If someone floods them, it's easier to tell of course.

If you want to protect individual traffic in a network you have to share access/are worried about passwords getting lost, the best solution is to go to WPA2 Enterprise with per-device credentials. On CCC-run hacker events they even use it for the "open" WLAN, and just accept any username and password.

Each client uses a different session encryption key but that is negotiated using the shared PSK.

I had to disable the ad blocker to get the nice web gl graphics.

From a front end perspective I think this it's awesome. No so sure about the content though.

I was looking at this and thought it looked nice but then my computer froze so maybe they overdid the 3D graphics a bit?

Lol I had the same thoughts

Try the Google cache, it now has the images but not the animations: http://webcache.googleusercontent.com/search?q=cache:http://...

the radio stuff is pretty wrong; one glaring one is that PSK is pretty much not used anymore, it's all OFDM.

OFDM is orthogonal to PSK (hah). PSK is a modulation--a way of representing bits on a carrier wave. Another type of modulation is QAM. OFDM is a way of combining multiple sub-carriers (each modulated with PSK or QAM) into one signal in order to deal with multi-path distortion: https://www.csie.ntu.edu.tw/~hsinmu/courses/_media/wn_11fall....

Some systems even combine PSK and ASK (like QAM) but asymmetrically (called APSK), obviously though you start getting into diminishing returns; noise limiting the gains. Of course you are limited by Shannon's law. But depending on the transmission medium lower energy signals can be more resistant to noise, so you can use either probabilistic or dynamic symbol selection to reduce the effect noise has on the signal and hence increase the available error free bandwidth (error free symbol rate)[0].

[0] https://en.wikipedia.org/wiki/Constellation_shaping ; check the references and external links for more in depth explanations.

Still, hatsunearu is correct. The page is grossly wrong, absolutely no modern WiFi modulation technique uses 8-PSK. Have a look at http://mcsindex.com/ 11a/11n/11ac all use BPSK, QPSK, 16-QAM, 64-QAM, and (for 11ac only) 256-QAM.

I know. And not a word on forward-error correction, trellis codes, automatic gain control, beam steering or MU-MIMO. Is this for babies? They don't even touch on low-noise amplifiers. Much less DAC and ADC performance. Where's the link a primer on FFT?

Can someone explain the natural resonance of walls of talked about with regards to 5Ghz?

Why did CSS stop working

Pretty sure that domain isn't owned by Verizon. Looks like an authorized seller creating content to try to build authority.

Looks like it is actually owned by Verizon, see email contact on WHOIS http://whois.domaintools.com/verizoninternet.com

wi-Fi is a best device! but is can work slow than wire. https://www.webhostingonedollar.com/ visit for getting more detail about hosting plan.

TL;DR: It doesn't.

website crashes chrome on windows after a few seconds.

In my experience it doesn't. My every attempt at improving this shit ends up in laying more wires everywhere.

Stopped reading at "Wi-Fi antennas send information". Antennas don't send anything, they just match the impedance of the feedline into that of the medium.

Reply on news.ycombinator.com