How to tune WIFI performance?

[kfmfe04]

Does anyone have any links to threads or WIKIs on how to tune WIFI performance for the N66U (both 2.4/5)?

Trying out a few tests on my LAN, I've noticed that WIFI to LAN (Ethernet) seems to be quite slower than WIFI to WAN. I suspect something might be wrong with my settings.

For example (latency/dl/ul):
WIFI to WAN 11ms 58.77MBps 14.17MBps
WIFI to LAN 2.35ms 37.3MBps 11.8MBps

 

[joegreat]

Does anyone have any links to threads or WIKIs on how to tune WIFI performance for the N66U (both 2.4/5)?

Trying out a few tests on my LAN, I've noticed that WIFI to LAN (Ethernet) seems to be quite slower than WIFI to WAN. I suspect something might be wrong with my settings.

For example (latency/dl/ul):
WIFI to WAN 11ms 58.77MBps 14.17MBps
WIFI to LAN 2.35ms 37.3MBps 11.8MBps

Hi,

There is no "one size fit's all" solution for the best WLAN settings.
It was discussed already quite often in the forum (see here and here for some examples).

In most cases the 2.4 GHz WLAN is quite overloaded with networks from the neighbors and you need carefully select the right channel (which is free or at least less busy).

But you really should start with fresh (clear NVRAM) and to only limited settings, to test the performance with defaults first.

With kind regards
Joe

 

[thiggins]

How are you measuring WiFi to WAN?

 

[Fraoch]

Seems like your LAN device has issues of its own...it should always be faster than WAN (unless you have 1 Gbps Internet!) I used iperf to test WLAN to LAN and got speeds similar to Tim's tests, the fastest was 96.9 Mbps simultaneous up-down on 5 GHz.

How are you testing it?

 

[kfmfe04]

I am using iperf:

On host A, iperf -s -B XXX.XXX.XXX.XXX
On host B, iperf -c XXX.XXX.XXX.XXX -d -t 60 -i 10

where XXX.XXX.XXX.XXX is the local IP of host A. Both machines are local.
I record the largest file transfer result from iperf.

• Host A is wired win8 running lubuntu 13.04 inside the latest vbox.
• Host B is wifi osx running lubuntu 13.04 inside the latest vbox.

I will try again when I get home later to confirm this weirdness.

I appreciate the sanity checks: would be nice to track this one down for large intranet file transfers or streaming.