WLAN Optimization broken?
I have compared the Omada Interference Detection tool with the Omada WLAN Optimization and something seems off.
Networks from a company next door show up with rather high 2.4 GHz interference on channels 1,6 and 11 (as to be expected).
Still, WLAN Optimization suggests I use channels 1,6 and 11.
Why? At least, choosing e.g. 3, 8 and 13 seem to be better?
Does the WLAN Optimization only optimize my WLAN without listening to traffic from neighboring networks?
My controller is an OC200 V1 Ver 6.2.0.17.
I have 4 EAP 653(EU)V1 units.
- Copy Link
- Subscribe
- Bookmark
- Report Inappropriate Content
@Commodus
I found a "vibe" solution, albeit not yet automated;
Feed the Interference Detection Tool data into Claude/DeepSeek/ChatGPT and define which APs are internal and which ones are external, and ask it to optimize internal AP channel distribution, power and bandwidth by forming a cost function based on rssi. You can position your APs in a coordinate system to take distances into the equation. Claude even models the shape of each band in the calculation.
It is an "NP_hard" problem and a hard force approach can be +1M cases to test, but Claude was pretty good at using heuristics and reasonable iteration.
As the output I get a table with AP channel, power and bandwidth. I could then further tweak it and ask it to prioritize the APs based on traffic and ask it to avoid the crowded channels (1,6,11) even if there was low interference reported, etc.
I was surprised how easy it was, but look out for hallucinations and ask control questions.
- Copy Link
- Report Inappropriate Content
Hi @Commodus
Thanks for posting your observation to TP-Link Business Forums.
The WLAN Optimization not just takes effect on the internal network, i.e., WLAN network, but also optimize the entire network environment.
The reason it would still choose 1, 6, 11 is that there are only these three channels that do not overlap with each other, thus it will only switch among the three channels.
- Copy Link
- Report Inappropriate Content
You can argue that channels 3,8,13 (in EU, not USA) are better choices than just try to reuse the often crowded 1,6,11.
- Copy Link
- Report Inappropriate Content
@Commodus
I found a "vibe" solution, albeit not yet automated;
Feed the Interference Detection Tool data into Claude/DeepSeek/ChatGPT and define which APs are internal and which ones are external, and ask it to optimize internal AP channel distribution, power and bandwidth by forming a cost function based on rssi. You can position your APs in a coordinate system to take distances into the equation. Claude even models the shape of each band in the calculation.
It is an "NP_hard" problem and a hard force approach can be +1M cases to test, but Claude was pretty good at using heuristics and reasonable iteration.
As the output I get a table with AP channel, power and bandwidth. I could then further tweak it and ask it to prioritize the APs based on traffic and ask it to avoid the crowded channels (1,6,11) even if there was low interference reported, etc.
I was surprised how easy it was, but look out for hallucinations and ask control questions.
- Copy Link
- Report Inappropriate Content
Hi @Commodus
I am wondering if you are saying this How to use WLAN optimization and RRM feature to optimize wireless networks | Omada Network Support
- Copy Link
- Report Inappropriate Content
Information
Helpful: 0
Views: 304
Replies: 4
Voters 0
No one has voted for it yet.