On a test AP to a CPE using V6.40.4 , I tried to disable 6,12,18Mbps but as screenshot revels 6Mbps is always used by Mikrotik,
Can someone give a good reason why the lowest rate cannot be disabled, this must be impacting on overall wireless performance if we cannot tweak setting fully?

Is any traffic going on that link? A low band will always be used when there is not much traffic on a link.

I see your screenshot saying "AP Client". Is your wireless interface is in the station mode? I believe the rate-related parameters only work in AP mode. In any case there's little sense in limiting the Basic rate set on a client as it cannot be narrower then on the AP.

I see your screenshot saying "AP Client". Is your wireless interface is in the station mode? I believe the rate-related parameters only work in AP mode. In any case there's little sense in limiting the Basic rate set on a client as it cannot be narrower then on the AP.

Both AP and CPE have identical wireless setting?

Is any traffic going on that link? A low band will always be used when there is not much traffic on a link.

Even doing a speedtest, 6Mbps is still present ?

Your speed is clearly higher, so why worry?
6Mbit will ALWAYS be in this list, because it is used for Management frame communication.

Is Management frame communication unique to Mikrotik and why couldn't this happen at higher speeds, can you explain a little or direct me to where I can do some more research on this,

As a WISP who is really trying to retain Mikrotik for wireless I have every reason to be concerned especially when I read the same chipset used by another wireless equipment supplier can give 180-200Mbps whereas Mikrotik struggle for 40/60.

So I am just wondering if Mikrotik wireless is suffering from "excessive code" used on ROS or put another way should Mikrotik be using a stripped down version of ROS for wireless?


Sent from my SM-A300FU using Tapatalk

Both AP and CPE have identical wireless setting?

Not necessarily. Standards say that wireless client should support all the basic rates that AP supports, otherwise it cannot (is not allowed to?) connect. That means changing the supported basic rate set on a client does not make much sense. If it is narrower on client than on AP, then connection should not be possible. And if it is narrower on AP than on client then the actual rate set in use will follow the AP's settings, no matter what settings client has.

Both AP and CPE have identical wireless setting?

Not necessarily. Standards say that wireless client should support all the basic rates that AP supports, otherwise it cannot (is not allowed to?) connect. That means changing the supported basic rate set on a client does not make much sense. If it is narrower on client than on AP, then connection should not be possible. And if it is narrower on AP than on client then the actual rate set in use will follow the AP's settings, no matter what settings client has.

We are not talking about "Standards" as used here is Mikrotik Propriety Wireless Protocol NV2 also customers have complained when data rates / MCS rates are different on a CPE compared with the AP?

nv2 will always use 6Mbit data rate for protocol communication.

used here is Mikrotik Propriety Wireless Protocol NV2

Do you think it was clear from your original post?
Don't answer...

nv2 will always use 6Mbit data rate for protocol communication.

Is it really a good idea using 6Mbps rate only for protocol communication in any high speed wireless connection.

@andriys
Oops - I edited topic title ?

nv2 will always use data rate for protocol communication.

Sincerely, Normis. Maybe you will try to increase the data rate from 6mbit for protocol comunication & in your laboratory conditions try 30-60 stations & saturation load. NV2 degrades the sector bandwidth if 30 or more stations are connected very strongly even if the signal is 50 dB or better.

What is the easiest way to monitor direct from Mikrotik router, of the total packets captured what is the % used for Management Protocol @ 6Mbps - packet analysis similar to Wireshark Protocol Hierarchy image below, capture done on a PC while streaming a video with an open Winbox session of the CPE.

Does anyone know if other wireless vendors
(a) use management protocols
(b) if so have they it set at 6Mbps

MTK wireless driver with AC not allow basic features like to read tx power, to make row frequency scan, to set power for different MCS and you ask to have full control of the data rate?

All we know that modulation switching take airtime.... I stopped to ask to mikrotik to have competitive wisp solutions and I switched to other vendor to stay on the market

Giuseppe

nv2 will always use 6Mbit data rate for protocol communication.

Does 802.11 use 6Mbit for Management Protocol
and
Does Nstreme use 6Mbit for Management Protocol

nv2 will always use 6Mbit data rate for protocol communication.

How much air time is used on an AP with 30 clients (NV2) when all clients have a 54Meg or faster connection speed for the "nv2 will always use 6Mbit data rate for protocol communication" ?

I would think if you have an active busy NV2 AP talking to dozens of NV2 clients during Internet peak usage hours (where the NV2 AP is aproaching saturation) and all clients have 100+ meg wireless link connections , that dropping way way way down to 6Mbit for "NV2 protocol communication" would be injecting pauses in a busy wireless network that could have a degrading effect on total wireless throughput.

I like to suggest that if possible & feasable , to have the "NV2 protocol communication" use what is at least in the Basic-Rates settings. Thus if everybody is running well at BASIC-Rates set to 24 Meg or even 54 Meg , then the time used for "NV2 protocol communication" would/could be much less and much faster - which should have an effect of faster AP-to-Client and Client-to-AP customer data stuff.

Sooo - the big question if if this is possible ... Just how much faster could an NV2 network run if "NV2 protocol communications" supported and worked with the configured Basic-Rates settings ???

North Idaho Tom Jones

nv2 will always use 6Mbit data rate for protocol communication.

How much air time is used on an AP with 30 clients (NV2) when all clients have a 54Meg or faster connection speed for the "nv2 will always use 6Mbit data rate for protocol communication" ?

I would think if you have an active busy NV2 AP talking to dozens of NV2 clients during Internet peak usage hours (where the NV2 AP is aproaching saturation) and all clients have 100+ meg wireless link connections , that dropping way way way down to 6Mbit for "NV2 protocol communication" would be injecting pauses in a busy wireless network that could have a degrading effect on total wireless throughput.

I like to suggest that if possible & feasable , to have the "NV2 protocol communication" use what is at least in the Basic-Rates settings. Thus if everybody is running well at BASIC-Rates set to 24 Meg or even 54 Meg , then the time used for "NV2 protocol communication" would/could be much less and much faster - which should have an effect of faster AP-to-Client and Client-to-AP customer data stuff.

Sooo - the big question if if this is possible ... Just how much faster could an NV2 network run if "NV2 protocol communications" supported and worked with the configured Basic-Rates settings ???

North Idaho Tom Jones

My concen and guess is that Mikrotik is fully aware of this potential wireless limitation and are unable to have mgt protocol use higher bit rates or any switching could impact on mgt protocol functionality : maybe just maybe Mikrotik have set ROS functionality a higher priority than wireless performance : all of this is of course speculation but I keep asking myself why does other vendor's with a similar chipsets and much simpler OS have better wireless performance ?

Sent from my SM-A300FU using Tapatalk

nv2 will always use 6Mbit data rate for protocol communication.

How much air time is used on an AP with 30 clients (NV2) when all clients have a 54Meg or faster connection speed for the "nv2 will always use 6Mbit data rate for protocol communication" ?

I would think if you have an active busy NV2 AP talking to dozens of NV2 clients during Internet peak usage hours (where the NV2 AP is aproaching saturation) and all clients have 100+ meg wireless link connections , that dropping way way way down to 6Mbit for "NV2 protocol communication" would be injecting pauses in a busy wireless network that could have a degrading effect on total wireless throughput.

I like to suggest that if possible & feasable , to have the "NV2 protocol communication" use what is at least in the Basic-Rates settings. Thus if everybody is running well at BASIC-Rates set to 24 Meg or even 54 Meg , then the time used for "NV2 protocol communication" would/could be much less and much faster - which should have an effect of faster AP-to-Client and Client-to-AP customer data stuff.

Sooo - the big question if if this is possible ... Just how much faster could an NV2 network run if "NV2 protocol communications" supported and worked with the configured Basic-Rates settings ???

North Idaho Tom Jones

My concen and guess is that Mikrotik is fully aware of this potential wireless limitation and are unable to have mgt protocol use higher bit rates or any switching could impact on mgt protocol functionality : maybe just maybe Mikrotik have set ROS functionality a higher priority than wireless performance : all of this is of course speculation but I keep asking myself why does other vendor's with a similar chipsets and much simpler OS have better wireless performance ?

Sent from my SM-A300FU using Tapatalk

It would be interesting if Mikrotik was to so this type of test:
- AP with many wireless clients
- Saturate the wireless traffic from the AP to all clients (then later reverse the direction and test again)
- Connect an oscilliscope to the AP antenna (or a seporate tuned antenna next-to/close-to the AP antenna). Or connect it to a test point to be able to see TX and RX traffic.
- Look at the oscilliscope for dead nothing-happening time periods.
- Then overlay those dead silent time-periods with a what is happening at that moment.
- Also ,,, overlay the time-periods of 6-Meg connection rates and see what is happening at that moment.

After the data is corelated , then have a wireless software programmer look at those time periods and attempt to make some wireless software changes in an attempt to reduce to a bare minimum all of the dead silent wireless TX & RX system where the wireless network is saturated.

As an old Electronics Engineer, I know getting a graph of TX & RX and dead silent periods should be fairly easy and simple to do.
However, It might require a good wireless software engineer to figure out what part of the wireless protocol/software was responsible for dead-air & 6-Meg rates and check/review/modify the wireless software code in attempts to acheive faster client throughput during wireless saturation.

Also - perform the same testing with a Point-to-Point (AP to Clinet) and perform the same procedures to improve customer wireless thorughput.

I am willing to bet , there is a superising amount of dead-air and 6-Meg connection rates , and other possible things that can be discovered and improved on to improbe throughput

North Idaho Tom Jones

nv2 will always use 6Mbit data rate for protocol communication.

How much air time is used on an AP with 30 clients (NV2) when all clients have a 54Meg or faster connection speed for the "nv2 will always use 6Mbit data rate for protocol communication" ?

I would think if you have an active busy NV2 AP talking to dozens of NV2 clients during Internet peak usage hours (where the NV2 AP is aproaching saturation) and all clients have 100+ meg wireless link connections , that dropping way way way down to 6Mbit for "NV2 protocol communication" would be injecting pauses in a busy wireless network that could have a degrading effect on total wireless throughput.

I like to suggest that if possible & feasable , to have the "NV2 protocol communication" use what is at least in the Basic-Rates settings. Thus if everybody is running well at BASIC-Rates set to 24 Meg or even 54 Meg , then the time used for "NV2 protocol communication" would/could be much less and much faster - which should have an effect of faster AP-to-Client and Client-to-AP customer data stuff.

Sooo - the big question if if this is possible ... Just how much faster could an NV2 network run if "NV2 protocol communications" supported and worked with the configured Basic-Rates settings ???

North Idaho Tom Jones

My concen and guess is that Mikrotik is fully aware of this potential wireless limitation and are unable to have mgt protocol use higher bit rates or any switching could impact on mgt protocol functionality : maybe just maybe Mikrotik have set ROS functionality a higher priority than wireless performance : all of this is of course speculation but I keep asking myself why does other vendor's with a similar chipsets and much simpler OS have better wireless performance ?

Sent from my SM-A300FU using Tapatalk

It would be interesting if Mikrotik was to so this type of test:
- AP with many wireless clients
- Saturate the wireless traffic from the AP to all clients (then later reverse the direction and test again)
- Connect an oscilliscope to the AP antenna (or a seporate tuned antenna next-to/close-to the AP antenna). Or connect it to a test point to be able to see TX and RX traffic.
- Look at the oscilliscope for dead nothing-happening time periods.
- Then overlay those dead silent time-periods with a what is happening at that moment.
- Also ,,, overlay the time-periods of 6-Meg connection rates and see what is happening at that moment.

After the data is corelated , then have a wireless software programmer look at those time periods and attempt to make some wireless software changes in an attempt to reduce to a bare minimum all of the dead silent wireless TX & RX system where the wireless network is saturated.

As an old Electronics Engineer, I know getting a graph of TX & RX and dead silent periods should be fairly easy and simple to do.
However, It might require a good wireless software engineer to figure out what part of the wireless protocol/software was responsible for dead-air & 6-Meg rates and check/review/modify the wireless software code in attempts to acheive faster client throughput during wireless saturation.

Also - perform the same testing with a Point-to-Point (AP to Clinet) and perform the same procedures to improve customer wireless thorughput.

I am willing to bet , there is a superising amount of dead-air and 6-Meg connection rates , and other possible things that can be discovered and improved on to improbe throughput

North Idaho Tom Jones

I would also ask does the 6 Mbit limit also apply to ethernet and fibre connections ?

Sent from my SM-A300FU using Tapatalk

n21roadie - Re: I would also ask does the 6 Mbit limit also apply to ethernet and fibre connections ?

I doubt it - they are totally different drivers & protocols.

However ... Ethernet and Fiber each also have their own additional things going on for initial link-connect-speeds , advertised rate capabilities , mdix , POE , along with other things.
Also - there is a big huge difference between 10/100 meg copper Ethernet cable wireing and Gig Ethernet wireing. Gig Ethernet actually is supposed to use 8-wires and 10/100 Meg Ethernet is really using only 4-wires in an Ethernet cable.
And - unlike typical Microwave (including all Mikrotik wireless cards) , Ethernet & Fiber can run Full-Duplex (send and receive at the same time). Microwave is always half-duplex , and can only send -or- receive one at time and never both at the same time. Sometimes it switches between send and receive so fast that it might behave like Full-Duplex but it is only Half-Duplex. Because wireless is only Half-Duplex , there are additional timing related protocols to make wireless work. A wireless AP can't properly send data to a Client while the Client is also sending to the AP at the exact same moment in time. This is where Mikrotik has created two additional wireless protocols (nstream an nv2). Both address different issues with the way wireless timing protocols operate and both have their own strengths and weaknesses. FYI - the Mikrotik nv2 protocol is actually is a properitery TDMA protocol used only by Mikrotik. The TDMA/nv2 timing is actually a hardware feature built into the Atheros wirless chipsets. Mikrotik nv2 is simply setting some registers/switches/configs in the Atheros chipset and then using some drivers to talk to the now-re-configured Atheros wireless chipset. TDMA has been around for a very long time - way way way longer than Mikrotik has been around. Mikrotik nv2 is just a Mikrotik protocol that configures/uses the TDMA capabilities of the wireless chipset.

With wireless , timing is critical to acheive throughput.
TDMA is a Half-Duplex timeslot protocol where all devices have defined periods in time where they can transmit
nstream is an AP polling protocol where the AP says to each client you can now transmit
802.11 wireless is more of a free-fall where it is possible to have multiple clients transmitting at the same time , which then can result in a small brief moment where data was lost. When data is lost using 802.11 wireless , then Random-Early-Detection (RED) kicks in and everything stops , waits a random amount of time and then begins again.
FYI - TCP-IP rate limiting / bandwidth queues actually uses the RED protocol to maintain communication speeds

North Idaho Tom Jones