Wi-Fi EasyMesh

https://www.wi-fi.org/discover-wi-fi/wi-fi-easymesh

I think openwrt is a good platform to implement Wi-Fi EasyMesh.

What's your opinion?

As my opinion, 66 pages of spec, including the arguable security implications of enabling TR-069 (remote management of CPE by ISPs) and "push-button configuration" (WPS) don't have me terribly excited.

It potentially could be implemented by a combination of user-space programs and driver modifications to implement both the logic of the protocol, as well as the new message types.

My opinion is that this is a forward-looking draft specification primarily intended to guide in-chip implementation for the next generations of devices.

Well, it seems that prplMesh has brought the idea into reality.

There is even prplMesh-openwrt and prplwrt !

Has anyone got experience on these?

I somehow fail to understand how this is different from a “normal” 802.11s mesh, other than it uses a similar method like WPS to add extra mesh points.

Could someone explain that in 3 easy sentences?

I am not sure, but I think you are right.
The main differences may be "easy" and "interoperable".

There are several mesh systems in the consumer market.
Some of them use 802.11s (Google Wifi?, Tenda Nova MW6?).
But there is no EASY way to make them INTEROPERABLE.

Hello.

Here is what I understood, happy to be corrected !
One of the difference is how the topology of the network works. EasyMesh relies on WDS and 4 address mode. A node of the network si both AP and STA and connects to the AP of another node. There is only one path between two points of the network, it's actually a tree. One of the nodes, the Controller, gathers metrics and capabilities of each AP and decides how AP are connected to each other. Commands are sent using a protocol based on 1905.1, specified by EasyMesh. An event can trigger a reconfiguration but the steering logic of APs is not part of the spec. Only actions by the controller are enabled by the protocol, but the decision logic is left to each vendor to innovate.

On the other hand 802.11s introduced the Mesh Point mode, which allows a node to maintain connectivity with all nearby nodes. There is no central controller which decides the topology of the network or data path. This is decentralized and part of 802.11s spec. And so, to my knowledge, there is no protocol to propagate a configuration on the network.

Hope this helps.

So, in theory, which delivers better performance? EasyMesh or 802.11s?

Effectively the same as it is the link throughput and latency that dominates in most situations.

Is code good enough to be deployed on live wifi network? Has anyone tested prplMesh in real world?

I wonder what the current state of prplMesh/EasyMesh for OpenWRT is. prplMesh seems to be actively maintained but unsure if there is any scope of "upstreaming" the implementation.

I wonder that, too.

It would be great to enable EasyMesh with routers/extenders that don't support OpenWRT but have support for EasyMesh.

Any news about this???

Cheers!

Hello!
I was interested to learn that EasyMesh and the 802.11s protocol are different entities. I had the idea to connect a router with OpenWrt and a Netis N3 router into one mesh network. This is important! There are routers on the market with OpenWrt support and routers without OpenWrt support, but with EasyMesh (for example Netis N3). Has anyone dealt with this problem? Are there any links to videos for beginners on this topic? Thank you and forgive me for the stupid questions.

Wired backhaul between whatevermesh and openwrt 11s or batman?

Netis N3 is from 2021 and there is absolutely no way it supports wifi alliance easymesh from 2024

Thanks for your answers, but...
You are probably confused, Netis produces two similar routers N3D, N3. Apparently, N3 it only produces for certain countries and it supports EasyMesh

Any web link to say "easy mesh setup using our app" is same wifi alliance easymesh?

You are a bit confused.

The WiFi Alliance EasyMesh is more like an automated WDS repeater setup with the same SSID/password and 802.11r (Fast Transition) protocol that they call Mesh.

Although there's an ongoing development by the PrplMesh Foundation, there are no easymesh packages for the OpenWRT we talk about in this forum.

There is, however, a package called luci-app-easymesh (https://github.com/ntlf9t/luci-app-easymesh), but all the info is in Chinese and you probably need to compile the package yourself and, from what I read using a transator, is not that automated and the roaming parameters must be set manually on the nodes.

More info about that package here.

@anon63541380 , I think it's not the same. They probably should mention EasyMesh with the proper capitalization and trademark sign, if it's compatible. For sure, TP-Link has a dedicated EasyMesh page with supported devices (https://www.tp-link.com/us/easymesh/), and I do have a Belkin RT3200 (same HW as Linksys E8450) that supports EasyMesh on the stock vendor firmware. I didn't try it because I flashed OpenWRT as soon as I got it and didn't had a second compatible device.

I wish a developer could compile and merge the required modules to install the PrplMesh implementation here, but I won't hold my breath.

Cheers!

Of course not! As for the Netis N3 routers, it is "EasyMesh™" as a registered trademark. I tested two routers, the above-mentioned Netis N3 and DIR-815/R4. DIR-815/R4 is a controller, Netis N3 is an agent. That's how
DIR-815/R4 (controller) <------- Wi-Fi ------> Netis N3 (agent)
Everything worked, the goal was achieved, the Mesh network worked between two routers from different manufacturers. I needed to achieve a seamless transition and it worked!

Nice, there is no open source agent software anyway.

Absolutely right! There is a huge confusion in terms and definitions due to their similarity. In my first question, I meant "EasyMesh™" as a registered trademark. It's my fault that I wrote it incorrectly, it should be like this "EasyMesh™" Thank you very much for the link, it will take me some time to study it.