Hello,
why mikrotik ROS no supported AC wave2 MU-MIMO?
What ist problem? Its driver problem or ROS kernel problem?

From my understanding its not supported on OS

But it would be nice to know if it's a WIP for ROS7 .

mikrotik team please answer!!!!

If you think they are listening to peoples pleas in the forum, you really need to get a life.
Bring a multi-million dollar contract to the table, it will happen in weeks.

Normis you here?

Why do you need Normis to tell you what you already know

i believe in Mikrotik

First you need hadware that supports it. Once that becomes available, you will suddenly find support in RouterOS...
I seem to remember that there was discussion about new hardware on the forum. The release of new hardware usually leaks because FCC has a website where all equipment that they certify can be found, including specifications, photos, etc. There are people who are watching that and post here when they find something.

i believe in Mikrotik

Agreed, Mikrotik is like any other religion,
a. it needs people to believe,
b. it wants your money (or possessions).

Seriously folks, MT is a business not a church LOL. They will make decisions that are deemed profitable.

Seriously folks, MT is a business not a church LOL. They will make decisions that are deemed profitable.

Obviously - and I sure hope they stay like this. The moment they start asking me to bow down and pray is the moment I will change brands.
So, the question really is: "When will wave2 MU-MIMO be a profitable decision?"

Normis you here?
Please answer!!

You should well know that Normis does not answer questions like that.

Mikrotik prefers people to buy more APs to handle more clients instead of giving them MU-MIMO, just saying

Comon Normis!!!

Wifi has its limitations mu-mimo might optimize the physics but there are hard limitations at frequencies used and the environments we use them in.
I would prefer clearer documentation, and stable systems and then the ability to tune the WIFI (without guesswork and head scratching AND THEN ......... add on capabilities.
Poorly documented mu-mimo that is unstable and not tunable has no interest from me.

I give up!

Wifi has its limitations mu-mimo might optimize the physics but there are hard limitations at frequencies used and the environments we use them in.
I would prefer clearer documentation, and stable systems and then the ability to tune the WIFI (without guesswork and head scratching AND THEN ......... add on capabilities.
Poorly documented mu-mimo that is unstable and not tunable has no interest from me.

anav, with all my respect, I think you could not be more off.
Modern Access Point Wifi (starting with 11ac Wave 2) and now Wifi 6 (Wifi 11ax) and Wifi 6E (6GHz) is nothing a single user or admin can really optimise anymore. I am not talking about PtoP or special outdoor usages.
For home/office/hospitality the algorithms are fast and deeply embedded and an "admin" will just screw up things.
All the roaming, steering is done on the fly, frequency optimisation, backhauling from Range Extender to CAP etc.. all automatic.
All Wifi clients do what they do and need to do, you can also not influence that anymore.
The Apples/Samsungs/Sony/Dells/HP's and name them all Wifi devices are tested and optimised to run with the latest and greatest SW seamlessly from
the top Wifi manufacturers. this is "built in".
You nor anyone else will be able to get more out of those Wifi system, you would only be able to reduce performance actually...

What you will be able to do is administrate the system: How many AP, which bands, connect AP and RE over Wifi or ethernet cable etc... but that's all.

I think Mikrotik also knows this and is doing something, now time will tell what and when things become available.

i believe in Mikrotik

I believe I need a different vendor for wireless.

Routers have been good to me.
Switches have burned me.
Wireless has dragged consistently. I invested a lot of time in caps-man only to have to abandon it for radio performance.

Stick with other vendors for switching and wireless and you and your clients will be a lot happier

I blame UniF--k... Had ubnt just stayed in the wireless space instead of putting out crap routers....

These 2 companies could have "stuck with what they do".

You nor anyone else will be able to get more out of those Wifi system, you would only be able to reduce performance actually...

There is a nice feature which is called "airtime fairness" which actually helps a lot and is used by a lot of vendors, not only the big ones

Why do we need Airtime Fairness?

Traditionally, your phones, laptops, pads or any other devices need to compete for the chance to transmit and receive data when they are connected to the same Wi-Fi signal. In this situation, once a slow transmitting device gets the chance, it will take longer time to send or receive the data. In the meanwhile, other faster devices must wait until the slow device finishes the transmitting process. Based on the above situation, you might want to cut down time given to legacy devices to allow faster devices download data for longer times. This will significantly increase overall capacity of the network. To achieve such objective Airtime Fairness is introduced.

How does Airtime Fairness work?

Airtime Fairness feature is based on TDMA technology, short for Time Division Multiple Access. It divides the Wi-Fi signal into many same time slots, and each Wi-Fi device takes turns to send or receive data from the Internet within its own time slot. In this way, the capacity and efficiency of Wi-Fi will be improved.



As shown above, without Airtime Fairness, those slow clients need more time to transfer the same data, which decreases Wi-Fi efficiency.

With two different speeds of clients connected to the same Wi-Fi, the test result below shows us the improvement of Airtime Fairness. When Airtime Fairness is disabled, the download speed of 802.11g client is 28.075Mbps while 802.11n client is 18.521Mbps. But when Airtime Fairness is enabled, the download speed of 802.11g client is 12.214 while 802.11n client is 116.538Mbps, which is highly improved. The overall throughput of Wi-Fi has tripled (from 46Mbps to 128Mbps).


(source: https://www.tp-link.com/ae/support/faq/2095/)

The nice thing about such protocols is that there are so many of them, and every manufacturer has their own ones.
That basically makes them mutually incompatible as well. Good for sales.
It would be helpful when everyone just implemented universal standards, and nu such nonsense as NSTREME, NV2, AIRMAX or other kludges...
But hey, then we could just buy equipment from any manufacturer, not what manufacturers desire, right?

I think Mikrotik is not using standard ath10k WiFi drivers like OpenWRT just cause they were stingy and put too little RAM on their APs (ath10k needs 256mb of RAM to work correctly).

I'm currently using OpenWRT on my AC2 with ath10k ct smallbuffers driver, much better coverage than ROS (higher TX power I guess) and MU-MIMO is also enabled. The port still needs some additional work till it gets into mainline branch https://github.com/openwrt/openwrt/pull/3037 but it's more promising than ROS 7 for home users xD

I'm currently using OpenWRT on my AC2 with ath10k ct smallbuffers driver, much better coverage than ROS (higher TX power I guess) and MU-MIMO is also enabled. The port still needs some additional work till it gets into mainline branch https://github.com/openwrt/openwrt/pull/3037 but it's more promising than ROS 7 for home users xD

You can adjust TX power on Ros.. isnt the same?

Sent from my SM-A705FN using Tapatalk

I'm currently using OpenWRT on my AC2 with ath10k ct smallbuffers driver

That is interesting. Did you perform some benchmarks to compare WiFi Speed between RouterOS and OpenWRT?
Is it possible to revert to RouterOS with included license?

much better coverage than ROS (higher TX power I guess)

For a fair comparison you should only compare legal TX power (depending on channel and sometimes country), not illegal settings that you may be able to do in an uncontrolled environment but you cannot reasonably expect from MikroTik.
You can calculate your max TX power by getting the max EIRP for your country for that channel (usually 20, 23 or 30dBm), subtract the antenna gain (given in product specs) and subtract another 3dB for 2-chain, 6dB for 4-chain.
Then you can set that power in your product under test (and verify that the MikroTik auto-calculated power is the same).

I'm currently using OpenWRT on my AC2 with ath10k ct smallbuffers driver, much better coverage than ROS (higher TX power I guess) and MU-MIMO is also enabled. The port still needs some additional work till it gets into mainline branch https://github.com/openwrt/openwrt/pull/3037 but it's more promising than ROS 7 for home users xD

You can adjust TX power on Ros.. isnt the same?

Sent from my SM-A705FN using Tapatalk

I have an US locked unit

I'm currently using OpenWRT on my AC2 with ath10k ct smallbuffers driver

That is interesting. Did you perform some benchmarks to compare WiFi Speed between RouterOS and OpenWRT?
Is it possible to revert to RouterOS with included license?

No yet, if you save your ROS license key then you should be able to use Netinstall and revert to ROS since OpenWRT doesn't modifies the bootloader.

I have a 50/50mbps and noticed that speed tests over 5ghz no longer slow down for a second on my S10e like in ROS, I think that's cause the AMSDU value is handled automatically by the driver instead of being a fixed value.

ROS it's good but Mikrotik should really consider dropping their actual driver in favor of ath10k in SOHO products, e.g. an indoor AP will never require to run PTP protocols like NV2 or nstreme instead they need better 802.11ac support.

So far the only downside I've found with OpenWRT are latencies, the default on demand governor takes too much time to ramp up the CPU frequencies and causes some spikes but changing it to performance gives a similar performance to ROS (the CPU is fixed to 716mhz).

ROS it's good but Mikrotik should really consider dropping their actual driver in favor of ath10k in SOHO products, e.g. an indoor AP will never require to run PTP protocols like NV2 or nstreme instead they need better 802.11ac support.

well, in theory it is possible to have 2 wireless packages. one called wireless-sp (or something similar) to add drivers for the non-mainstream things.
but again, two source trees to maintain/troubleshoot/test. plus you might don't know what you would lose. i like capsman for example. and i am not quite sure whether it would work with the default atl10k driver. it's hard to make good decisions without seeing the entire big picture.

RFC1925 (4)

I'm currently using OpenWRT on my AC2 with ath10k ct smallbuffers driver, much better coverage than ROS (higher TX power I guess) and MU-MIMO is also enabled. The port still needs some additional work till it gets into mainline branch https://github.com/openwrt/openwrt/pull/3037 but it's more promising than ROS 7 for home users xD

Interesting.

• Will it work on MikroTik cAP ac, aswell? (similar hardware, same chipset)

• Does it support MU-MIMO?

• Does it support IEEE 802.11k?

• Does it support IEEE 802.11v?

• Does it support IEEE 802.11r?

• Does it support band steering?

• Does it support airtime fairness feature?