A lot of Wi-Fi systems are slow because there is a lack of understanding about the bit rate overhead. In this video I give a demonstration of some of the most common causes.
I actually tried doing this recently, but I have found that a lot of devices go into a power saving mode and go down to a 6Mbps bitrate quite often. disabling the low bitrates has then caused connectivity problems, as clients then just disconnect. So there is still a balance. And changing to a min of 6Mbps is still quite a drastic increase in spare air time.
For anyone who didn't understand, the big takeaway here is that if you want fast Wi-Fi, you need to put it on a treadmill. And the more of them you put on the treadmill, the faster it will go! 😉
Absolutely excellent demonstration! I already knew the basics about this however your demo with the SDR really helped show exactly what is happening on the air!
And thanks for the WiFi consortium to define channel 1 to 11 (13,14) even only 3 are usable anyway, so your neighbours choose something in between to destroy a lot of otherwise usable airspace.
Interesting topic. How do more modern technologies handle this? Few people use .11G any more. How does 802.11AC or AX handle these things - and in particular - interference from outside AP's?
If your wifi is slow, check if you have Bluetooth speakers interfering, cordless DECT telephones, RF gadgets, your neighbors, or even just your old microwave oven that started leaking. I can knock off my phone from my weak 2.4 GHz WiFi in the kitchen by placing the phone next to the microwave and switch it on
Next to the microwave is expected. The shielding is never perfect but it should attenuate it enough that anything a reasonable distance away has minimal interference.
"You wouldn't have 5 APs sitting on top of each other". The thing the comes to mind is a "trade show environment". This could be a good start when it comes to specifying allowable parameters in that setting, which certainly different than what a hospital would require.
Careful with low power IoT devices, they're normally good if you move bacons from 1 to 5.5Mbps, I find if you set your Internet of Turd SSID to 11Mbps some devices stop responding. 6 Mbps (12 x500kbps units) Basic Rate value represent as 10001100 7th bit =1 in your beacon and nothing left for the mcast dns, try in the next frame pls :-)
Sadly, Wi-Fi is way to easy for any muppet to throw out there without a clue and flood the spectrums. The most common issue I get is over saturation of WAP's, all too close to each other and many are even big-name brands that many fall for the $$$ means it is better opinion. Many devices have big RF power but small antenna systems which result in screaming outputs but low sensitivity to hear/listen for weak signal devices trying to get back. Aruba and Ruckus good for this. I use a lot of UniFi and see the noise created by itself as well as the flood of other junk (ISP supplied cheap modem/routers are a good one on 40/80-160MHz bandwidths for 2.4/5GHz) but I have seen how a busy network can be destroyed in minutes by too many WAP's all screaming their heads off. Mining and Construction camp sites are my bigger clients as well as schools. Great video BTW.
2.4 ghz doesn't really have more than 3 usable channels. The band width of those channels overlap, so they have to be multiple channels over to not overlap at all. It is *good* when people only use channels 1, 6 and 11, because APs in the same channel can take turns talking while APs in different channels interpret each other as interferences and have no choice but to re-transmit in the hopes that it will go through..
Using only three channels in 20MHz chunks is actually preferred, as long as they're on channels 3,7,11 here in America. They'll magically cooperate to all work separately as well as possible, as opposed to destructively interfering like a bunch of 40MHz Eero devices fighting to the death over a 30MHz chunk of airspace would do.
It may not matter in a single home, but in an apartment building with neighbors above, below and on every side with their networks on high power it's a nightmare. This was a GREAT video. Thank you!
I know what I'm checking tomorrow at work! Paul thanks for the videos, found your Channel today via the Tesla video and now digging into your back catalogue, all very interesting! Looks like your subcount will be going up this week as the algo is promoting you. all the best
came here to say this, I work in networking but I am fascinating by wi-fi, despite the space age RF tool, the rest of it was really simple to follow, top work!
Not sure I'm just seeing coincidence when there is none, but when devices like mobile phones or ESP32 devices on my WiFi network go into low power (PSM?) mode, they seem to talk slower WiFi speeds. I wonder if setting a higher minimum bitrate would interfere with that? I can't seem to reach my ESP32 devices when I set the minimum above 6Mbps. I'm using Tasmota and this might be a limitation of that or of the hardware itself.
cranking up the power is unfortunately often necessary due to mobile devices and other weird stuff. channel reuse is a pain no matter what power level you use. Wi-Fi 6E will save us.
Increasing access point Tx power too much can cause issues as well, depending on the Tx power limits of the client devices. If the access point Tx power is significantly higher than a client’s, then the client’s transmissions may not be able to reliably received by the access point for clients near the edges of the AP’s coverage. This could also mean other clients closer to the access point don’t known when the far client is transmitting, resulting in increased collisions, causing retransmissions and resulting in reduced performance.
As long as I've explained this to customers over the years I've never bothered to view this over the spectrum like this... I've bookmarked this one and will be forwarding it on to those that beed a little more convincing. Awesome stuff Paul!
Very first thing I do before deploying an AP is pull up a WiFi analyzer to find the channels not in use. After that it is all easy. Great video and learned allot that I didn't know. You get a subscribe and a thumbs up!
Power is becoming a major issue at home too. So many people don't get more bars doesn't mean faster and you can get cheap extra high power mesh junk online that gives you more bars so people think its better. Someone in my area (kinda rural, 100+ feet between houses minimum) got one and they are stronger RSSI in my own livingroom than the APs in my own house next to me...I actually had to redo my channels to work around their interference.
I bought a nighthawk mesh system and it works super well, but I did give my TV its own Moca adapter because it's in a weak spot and buffers with >50Mbps media. And I'm sure this video applies, too.
If I'm correct, beacons are sent on the main 20MHz channel and the second or even third and fourth channel will be added for a client just for data. Some APs let you configure if you want base channel +20MHz or base channel -20MHz for the second channel.
@@TallPaulTech Just the fact you don't have a obnoxious intro on every video, no Warthunder ad in the middle and "like&subscribe" ending make them a lot more enjoyable 😄
Probably not. Most are shit, especially home ones. The ones that do have the option still need a person who knows what it means to make the right settings.
Well using AM radio in aircraft may have so rational. But the root cause is just inability to evolve. Most Aircrafts rules are just the same. Using Q codes is on big example. The only clever move would be using digital communication and reducing radio chat by sending number and instructions directly to screens.
Huh the first thing I look at is the type band, PHY rate, MIMO and number of spatial channels and so on. If all is good there, the rest typically falls in place...
How the hell do you even fit 90 devices within 2m of each other? Turn the bit rates up, and turn the power down. Add a couple more APs on DIFFERENT channels, not right on top of the clients (as they'll hear adjacent channels anyway being that close)
@@TallPaulTech A Canadian online acquaintance of mine does IOT systems development and can easily have up to 200 devices in a room (his workshop). I wouldn't say it's impossible but the tuning for such density is going to be quite insane, I'm not even sure there are APs that can drop the transmit power and cell sizes that small, not to mention such devices typically don't have (assisted) roaming capabilities.
Hi Paul, could you do a video showing the signal attenuation through different materials wall , wood , glass etc... and how should one calculate the number of APs based on that. Thanks
Good idea, but it could be a bit tricky, as I don't have a lab environment for that, and bits of RF would bounce around all over the place. I'll see if I can think of anything.
Do some simple benchmarking in your own circumstances. Building materials, typical wall thicknesses etc. Place an AP in a room, configure an ESSID at 2.4 GHz and another one at 5 GHz, and walk the rooms and corridors around with a scanner of some sort. The Inssider software by Metageek in some past version was my favourite simple scanning tool for Windows, up until a certain point where it became trialware or something. At the moment they require you to register to use the tool. In Linux there should be something similar... I can see LinSSID. After you get the hang of how far the signal reaches while remaining practically useable, you'll be able to place the AP's pretty easily, given a rough floor plan. For example, in the building where I work, 5 GHz gets through a 10cm brick wall, but not through a floor/ceiling, and 2.4 GHz gets through 2 walls or 1 floor/ceiling. Which means that I'd need about twice the number of 5 GHz AP's, compared to 2.4 GHz AP's :-) So in my case I decided that 5 GHz was a treat/bonus and didn't have to work everywhere. And, I allocate 2.4 GHz such that no two adjacent AP's share a channel. I enjoy the luxury of being the supreme master of the WiFi spectrum in our building, which does help :-) and which typically is not the case in residential blocks...
You should not use overlapping channels like 1 and 2 or 6 and 7 because devices may not "hear" each other transmit on opposite end of their respective channels.
Under LuCI: - Network > Wireless - Edit your SSID - Under 'General setup' uncheck 'Allow legacy rates' - Under Advanced Settings, you can configure the basic rates under coverage cell density: Normal, Disabled - Allows all rates High - Disallows
Or its because the persons Router itself or Configuration sucks. And is it not true that many routers dont give you the option to change settings like that? So isnt this more in line with commercial hardware settings than residential?
Am I correct in that the Transmit Rates mentioned on the GUI of Aruba is the same as OpenWrt 22.03.5 legacy_rates = no (Disallow legacy 802.11b data rates)?
After watching this, I checked all my APs to see how they were set, and on my OpenWRT routers that was my best guess. At least I was able to go onto my Engenius AP's and set the minimum to 11 Mbps as suggested, although I should really do a site survey and see if I have any clients using the old 802.11 versions in which case disabling the legacy 802.11b entirely is probably best. I have a HackRF and I knew it could pick up the 2.4GHz spectrum (I've fiddled with that) but it was definitely interesting to see it put to this use, and now I'm curious what kinds of interference I could find that seems to occasionally slow things down. Nice tip there.
I want to know how it is that DJI is able to get 2.4/5 GHz to support controlling a drone with streaming video over 8 miles but I can’t get a decent 5 Ghz signal from the room next to me. It goes to show it’s possible. Either the 802.11 standard is crap or there’s lots of really bad implementations
🎯 Key Takeaways for quick navigation: 00:00 📡 Low bit rates can significantly impact Wi-Fi network performance. 03:23 📈 Increasing bit rates on Wi-Fi SSIDs can improve network performance. 08:02 🚧 Adding multiple SSIDs with low bit rates can congest and slow down a Wi-Fi network. 12:58 ⚙️ Increasing AP power can lead to more collisions and worsen network performance. Made with HARPA AI
Yes, GR8T INFOtainment. It was a bit fast paced for my gray matter, but that's a "me" thing, not you... I firmly got the gest of it though. Like String theory j8Oo
What are your thoughts on using ad-hoc mode for home network to reduce RF pollution, so to avoid continuous beacon transmissions? It is also possible to use on some routers...
SSID != Radio. all of the ssids come from the same radio chip so obviously they run on the same frequency as otherwise the radio would have to switch all the time loosing quite alot of throughput
Thanks! That's right, you did. But I didn't understand the significance of it until I'd watched the whole video, and by that time, I'd forgotten you'd said it! :)@@TallPaulTech
AT&T’s uverse routers still use Wi-Fi 4. Google fiber install tech here. I even seen those routers used in conjunction with AT&T fiber many times. For some reason they only provide their Wi-Fi 6 equipment if they can bring the fiber inside the home, if not your stuck with their aging 2007 uverse equipment. Supposedly customers have told me that the fiber jacks are locked to the uverse routers so you gotta use their flaky bridge mode and hope a firmware update doesn’t factory reset the equipment.
@@francocastilloAR as long as customers aren’t trying to watch RU-vid or Netflix -45+db, 25+ft or a wall and a kitchen/bathroom/utility wall away from the router. Old houses has some terrible layouts for isp routers