Why my iPhones has much faster speed than other devices

I have Cudy X6 device running OpenWrt 22.03.02. Just clean one without any package installed.
It configured to run only 5GHz wireless that configured to running on channel 100 (other channels 36, 56 and 157 have the same results), in AC mode and 80MHz.

The router is standing directly on my computer than is standing on my table.

Another computer is standing next floor, directly above me (about 3 meters). It has 2 Wi-Fi cards.

  1. Intel Dual Band Wireless-AC 3168 with single external antenna.
  2. Realtek 8812BN Wireless LAN 802.AC USB NIC with 2 external antennas.

When I do speed test (on speedtest.net) I'm getting around 270-280 megabits. When I do speed test, I disable one adapter and enable another). I also run iperf in 5 and 8 parallel clients and got about the same result.

I tried notebook in the same position and results are little bit better around 285-290.

But when I run iPhone 8 Plus and iPhone XR at the same position as these external antennas are, they are able to get 430 megabits.

I always thought that iPhones are not that great at network speeds because they are relatively small and have power limitations.

I just thinking maybe there are some settings that can improve compatibility and speed of my router with these network adapters? I checked all settings on adapters in Windows and none of them improve speeds.

I tried to replace router with older TP-Link TL-WDR4310 running OpenWrt 18.06 and place it at exactly the same spot. All network adapters and iPhones have the same speed around 160 megabits.

I'm just thinking perhaps there is some settings that right now are in favor of iPhone network chips, but perhaps I can dial them in favor of other network cards?

Could be that the iPhone support more parallel spatial streams and hence allow higher theoretical throughput. Have a look at the "Associated Stations" list on LuCI's Status->Overview page, it will report information for each connected station/device, maybe the iphones report different information compared to the rest.

Thank you for an idea.

To me it looks about the same. Sometimes numbers on iPhone are higher, sometimes on network adapter. iPhone is on the bottom

SNR reported is quite different - if same location, perhaps antenna capabilities?

Now iPhone is on top and different network adapter

SNR reported is quite different - if same location, perhaps antenna capabilities?

Yes, I agree, but I don't think that notebook and 2 different network adapters on PC are having much worse antenna capabilities.

Is there any way to see them?

1 Like

I have not done such characterization.

Some help may be found here:

I would suggest assessing MIMO and connection speed vs both SNR and single device use.

1 Like

i have a 15 year old laptop with a draft-n wifi card that connects at 450 and a 8 year old wifi card draft-ac that connects at 866, it's not about beamforming and other stuff, it's simply about having multiple antennas with actuall mimo streams, and cheap noname stuff usually only has one-two.

LE: now i see it's your router that is more obscure, you can try to contact it's maintainer, probably something is not standard in it.

LE2: try to play with Fragmentation Threshhold, Beacon/DTIM interval, Distance, Coverage cell density under Advanced settings

1 Like

Anything I have to look for? I checked related article and I tried to use my PC as mobile spot (and switched off router). I got 190 megabits because Windows only allows 20 MHz channel. I would imaging that if I would be able to use 80Mhz, it would get close to 4 times faster speeds.

So, looks like I have the same problem but there is no solution there :frowning:

Thank you for idea. I did some research and I found this:
Coverage cell density looks like does nothing at 5GHz

Fragmentation Threshhold - it looks like it is used when there are multiple APs on the same channel. I'm sitting on channel 100 and there are no APs on the same channel (or close)

Beacon/DTIM interval - there are a lot of different kind of information, but it also looks like it works only with multiple clients. I have only one client.

I couldn't find any information about distance, I will play with it later.

Not wanting to sound pessimistic, but there is possibility that this indeed has no solution…

I believe key would be to establish if your ‘slow device’ can achieve maximum throughput (as the only device on the wireless network)

If it does, it would seem to point to something like the beamforming algorithm (on the router?) that may not be optimal with multiple devices present.
If it doesn’t, there most likely is something wrong at a lower level (probably on your ‘slow device’)…

Ok, I found reason. I did research and found that Intel card is actually 1x1. And my understanding that it is normal kind of speed for 1x1.

Then I did research for Realtek 8812BU and found that it is 2x2. But speed is exactly the same as 1x1. After poking around, I decided to update Windows drivers for it. It didn't update anything, and version of driver stayed the same, but speed magically jumped to 400-430. iperf3 constantly showing 450. Perhaps driver setup triggered something in network card itself and perhaps enabled 2x2. Or perhaps setup changed driver settings.

This topic was automatically closed 10 days after the last reply. New replies are no longer allowed.