802.11n

Mon May 25, 2009 10:39 am

I would be great if the first people to test the new 802.11n support in RouterOS v4 beta 3 would post their experiences here in this thread.

Anyone who will post some 802.11n wiki tutorial with images and setup configurations that work with good speeds, will get a Level6 license from me

angboontiong · Mon May 25, 2009 10:41 am

This mean the 11n still have no yet fully test?
If then, it's take risk to implement it into the live network...

Mon May 25, 2009 10:48 am

of course it's tested, but this is still BETA I'd like to hear success stories from users.

erny · Mon May 25, 2009 12:27 pm

it would be interesting to hear, are there any tests for backhaul links 5+ km range in 802.11n mode and what speeds are(could be acheived) i.e. with 2 dish antenas on both sides.

angboontiong · Mon May 25, 2009 1:11 pm

Hope can heard the feedback very soon....

tyronzn · Mon May 25, 2009 4:12 pm

I wouldnt exactly call this a success story,but maybe this ought to answer alot of people's questions with regards to the SR71A made by Ubiquiti and RouterOS.This morning i upgraded my two RB433AH routerboards to the lates version 4 beta version as well as upgraded the BIOS,clicked "update license key" and pronto,after a quick reboot Mikrotik detected the two cards.Enabled them and got them wirelessly associated to each other.Although i only got 35mbps through them on a throughput test,i didnt exactly spend much more time trying to fine tune it as i ran out of time.But both boards were simply sitting next to eachother on the workbench and noise levels were quite high which affected my CCQ,which in turn affects the total throughput.Perhaps when i get some time i'll do an outdoor link and post test results.

titius · Mon May 25, 2009 4:44 pm

You used two antennas on each card?

tyronzn · Mon May 25, 2009 5:48 pm

hello

yes i used two antennas on each card,i did an outdoor test over a couple hundred meters using a dual polarized antenna with RB433AH's either side using the SR71A's.I also had two pc's connected to each board on both ends and used the mikrotik bandwidth test to do the tests through the boards for accurate results.I Suspect i might have been able to get faster speeds as the 433AH's only have 10/100 ethernet ports as compared to the RB600A's with Gigabit.Despite that i managed to consistently get 95.7Mbps without any drops.In my opinion i think the new R52N would perform a bit better though

omega-00 · Mon May 25, 2009 7:27 pm

I've been doing some reading up on 802.11n now that it's supported and testing with 2 x TP-Link 802.11n cards I'd purchased a while back.
I've also posted some info on the tests I've been doing here: http://www.air-stream.org.au/blog/omega ... -and-cards
---
Mikrotik Controllable Options

HT Extention Channel - none, above control, below control
User selects a control (primary) channel (frequency) and sets if the bundled(bonded?) channel is above or below the selected channel.

HT AMPDU Priorities
AMPDU allows bursting 802.11 frames up to 64KB and is performed in the hardware. Not quite sure what these settings do yet

HT Guard Interval - any, long
Short GI is 400 nanoseconds vs. the traditional GI of 800 nanoseconds. Short GI reduces the symbol time from 4 microseconds to 3.6 microseconds to improve throughput by 10%.

HT MCS - Currently unchangable
MCS selects, based on RF channel conditions, the best combination of 8 data rates, bonded channels, multiple spatial streams, different guard intervals and modulation types (noting that 802.11 b/g adapts to channel conditions by selecting the highest of 12 possible rates from 1 to 54 Mbps)

Antenna Selection - Not sure if this should be enabled for use with 802.11n Given the Tx/Rx Chain settings?

HT TX/RX Settings - I'm still not entirely sure how the Tx and Rx chain pairing works, but I know that the card I have has 2 antenna sockets although I see only a small increase between having 1 or more selected.

----------------------
Improvements we might be seeing available in Router OS:
Preamble Mode: long, short, both, greenfield (new)
Greenfield preamble cannot be interpreted by legacy stations. It is shorter than the mixed mode or legacy mode preamble and improves efficiency of the 802.11n networks with no legacy devices.
----------------------

Would any mikrotik reps care to correct me here or add any further comments?

Pics of my testing

titius · Tue May 26, 2009 9:11 am

well these results are bad. 50 Mbps? no good for N

WispZA · Tue May 26, 2009 10:26 am

Okay...

After a day or so of messing about I've gotten the following results...(thanks to omega-00 for the excellent information)

[w/ conn. track]
TCP: 60.6MBps (peak) 60.1MBps (average) Very stable (w/ conn. track).
UDP: 179.8MBps (peak) Very unstable.
Ping Times - 1ms/2ms/22ms (min,avg,max)

[wo/ conn. track]
TCP: 64.5MBps (peak) 63.2MBps (average) Very stable (wo/ conn. track).
UDP: 196.7MBps (peak) approx. 165MBps (average) Stable.
CPU Usage approx 30-35% on Master Draft-N unit (bridge unit).
Ping Times - 1ms/1ms/8ms (min,avg,max)

Using 2x RB600A boards and TP-Link Draft-N cards.
Using 2x RB600A boards to generate traffic.
Very noisy environment. Scanning detects 5x APs in the building. One AP running 100mW about 10 inches from the one test board. Turning all this off makes no difference.

Channel: 2412 (other frequencies perform more poorly)
AMPDU: All selected on both units (definite bandwidth improvement)
HT Extension Channel: Above Control (definite bandwidth improvement)
HT TX/RX Chains: All selected on both units (definite bandwidth improvement).

These three changes upped the UDP bandwidth from 95MBps peak to 179.8MBps peak on UDP and from about 44MBps to 60MBps on TCP (stable). Disabling connection tracking boosts the TCP average by about 3MBps and makes UDP more stable with the peak improving by about 20MBps.

Tried various other settings, but none made a difference (except to possibly worsen it)

Will keep fiddling.

Edits in RED! + Images

WispZA

jsuter · Tue May 26, 2009 1:23 pm

A> does nstream (single and/or dual) work with the 11n adapters yet?

B> Does the MIMO work in b/g mode (support 2x2? 3x3?)?

C> Can you control the channel usage (ie - using it on sectored towers, I'd rather avoid as much crosstalk as possible)

D> 5/10 mhz rates? Turbo-a/g compatibility? Is there a such creature as HT5 and HT10?

E> Can I use a PCI Express type card (ie - in a x86 motherboard's PCI-Ex 1x slot?) with MT?

F> My logic on MIMO applications is the antennas should be in the same polarity (idea being that you're more likely to get a narrow obstruction [tree limb, light pole, whatever] than a cleanly spun signal). Any input on this?

I'm hoping to test b/g or g-only MIMO on customers that have issues with low data rates due to tree multipath.

Tue May 26, 2009 1:29 pm

Anyone who will post some 802.11n wiki tutorial with images and setup configurations that work with good speeds, will get a Level6 license from me

angboontiong · Tue May 26, 2009 1:52 pm

can post the speed test with the ping times as well?

omega-00 · Tue May 26, 2009 6:47 pm

Anyone who will post some 802.11n wiki tutorial with images and setup configurations that work with good speeds, will get a Level6 license from me

Stage 1 of a comprehensive guide to 802.11n on RouterOS completed.

http://wiki.mikrotik.com/wiki/802.11n_Setup_Guide

Seeing as its 3:30am here, I think its time for bed and more testing tomorrow

gustkiller · Tue May 26, 2009 8:46 pm

does 802.11n works on the 5ghz band

Wed May 27, 2009 8:40 am

does 802.11n works on the 5ghz band

yes, sure. read the specs of the cards:
http://routerboard.com/pdf/R52n.pdf
http://routerboard.com/pdf/R2n.pdf

Wed May 27, 2009 8:43 am

Anyone who will post some 802.11n wiki tutorial with images and setup configurations that work with good speeds, will get a Level6 license from me
Stage 1 of a comprehensive guide to 802.11n on RouterOS completed.

http://wiki.mikrotik.com/wiki/802.11n_Setup_Guide

Seeing as its 3:30am here, I think its time for bed and more testing tomorrow

I moved your supported card list to the usual page:
http://wiki.mikrotik.com/wiki/Supported ... I_Adapters

omega-00 · Wed May 27, 2009 12:24 pm

Cool, didn't think of that. Thanks

Wed May 27, 2009 12:26 pm

Uldis also updated some missing info in your tutorial, more specifically the Chains. I see that you used an old generation card with three chains, new ones use only two to get the same performance. Chains - Antennas.

omega-00 · Wed May 27, 2009 1:11 pm

All good, I'm currently waiting for my local supplier to get in the mikrotik cards so I can test more, glad to see that the information I've collected so far is correct and useful

Are you guys collecting supout.rif's for devices running with 802.11n? If not currently, do you want any? Or only if we see any problems.

Either way I've noticed a couple of issues:

The signal strength reporting for the HT speed rates fluctuates wildly in winbox, this may be a problem with my setup but as noted I'll check further when I get my R52N's

2jarek · Wed May 27, 2009 1:32 pm

hello

yes i used two antennas on each card,i did an outdoor test over a couple hundred meters using a dual polarized antenna with RB433AH's either side using the SR71A's.I also had two pc's connected to each board on both ends and used the mikrotik bandwidth test to do the tests through the boards for accurate results.I Suspect i might have been able to get faster speeds as the 433AH's only have 10/100 ethernet ports as compared to the RB600A's with Gigabit.Despite that i managed to consistently get 95.7Mbps without any drops.In my opinion i think the new R52N would perform a bit better though

Tyronzn
95.7Mbps UDP ? TCP ? packet size ?

triztrz · Wed May 27, 2009 2:17 pm

the wlan dont detected with TP Link TL-WN851N
x86 with Mikrotik v4.

anyone can help?

Wed May 27, 2009 2:20 pm

did you update license as per instructions ?

Wed May 27, 2009 3:00 pm

The signal reading are combined ones. In future we will improve the signal readings (maybe by adding signal for each chain). If you want to align the antennas, then you can do that by aligning each individual chain.

Stanislav Hrncir · Wed May 27, 2009 5:18 pm

... and what about users with RB 600, default os V3 and "upgradable to: V3.x" ?

wavemich · Wed May 27, 2009 11:10 pm

Does each chain (or antenna port) using .11n use a seperate 20Mhz channel or do they all share the same 20Mhz and transmit at the same time and recieve at the same time? I've not found any good documentation to explain that.

InoX · Wed May 27, 2009 11:22 pm

that's the funny thing...you need another RB600 or another license

kapulan · Wed May 27, 2009 11:24 pm

The "n"card support nstream & 40mhz channel separation?

niksoon · Thu May 28, 2009 12:08 am

802.11n tested on RB433AH and R52n

niksoon · Thu May 28, 2009 12:14 am

802.11n tested on RB433AH and R52n

in office on RTS-Computers/Bulgaria
from Nicky Krastev and Vladimir

niksoon · Thu May 28, 2009 12:31 am

In very soon will testing new 802.11n on distance of 20 kilometers and
will provide demo access .

omega-00 · Thu May 28, 2009 8:06 am

Got another test from mine as well, first one had some spare antennas of different gain db, now changed to all the same.

Updated the wiki guide to include this image.

magnavox · Thu May 28, 2009 8:43 am

Great!

Can you post more details on your hardware configuration?
Antenna, distances, signal strengt..

magnavox · Thu May 28, 2009 8:48 am

Got another test from mine as well, first one had some spare antennas of different signal strengths, now changed to all the same.

Updated the wiki guide to include this image.

Can you post TCP speedtest?

kapulan · Thu May 28, 2009 9:28 am

802.11n tested on RB433AH and R52n

Great speed!You can test two side the x86 pc based system?

Thu May 28, 2009 9:43 am

... and what about users with RB 600, default os V3 and "upgradable to: V3.x" ?

Please upgrade to v4.0beta3 and go to the sysltem license using winbox and update the license, it will give you a key with the correct "upgradable till"

Thu May 28, 2009 9:45 am

Does each chain (or antenna port) using .11n use a seperate 20Mhz channel or do they all share the same 20Mhz and transmit at the same time and recieve at the same time? I've not found any good documentation to explain that.

they all share those 20mhz and transmit and receive at the same time.

Thu May 28, 2009 9:56 am

The "n"card support nstream & 40mhz channel separation?

currently it doesn't support Nstreme, we will try to add that later.
What did you mean about 40mhz channel seperation? You ment turbo mode (40mhz channel)? N card doesn't support that, but it can use an extension channel (another 20mhz channel) above or below the center freqeuncy.

niksoon · Thu May 28, 2009 11:56 am

802.11n tested on RB433AH and R52n
Great speed!You can test two side the x86 pc based system?

Thanks,
Yes, I will made test x86 pc based system(Gbit ethernet) with two n-radios and bonding in each side, also I will put two dual polarization antennas per each side.

magnavox · Thu May 28, 2009 3:46 pm

802.11n tested on RB433AH and R52n
Great speed!You can test two side the x86 pc based system?
Thanks,
Yes, I will made test x86 pc based system(Gbit ethernet) with two n-radios and bonding in each side, also I will put two dual polarization antennas per each side.

What's the distance?
Can you post a test for single radio (not "bonded") ?

richinuk · Thu May 28, 2009 6:07 pm

Does each chain (or antenna port) using .11n use a seperate 20Mhz channel or do they all share the same 20Mhz and transmit at the same time and recieve at the same time? I've not found any good documentation to explain that.
they all share those 20mhz and transmit and receive at the same time.

Hi Uldis,

Are you saying it's full duplex?

Rich

wavemich · Thu May 28, 2009 6:52 pm

.11n is definitely half-duplex unless you incorporate Nstreme2 in some fashion. Which leads to the next question. Under Nstreme2 -> HT, what is the difference between 'single', 'double', and 'both'?

Will these .11n cards ever be capable of supporting full duplex capability using just one card under Nstreme2?

-jon

triztrz · Thu May 28, 2009 8:22 pm

the wlan dont detected with TP Link TL-WN851N
x86 with Mikrotik v4.
anyone can help?

did you update license as per instructions ?

upgrade the version from v3.24 to v4.0beta3
and i already update the key.. follow instruction from wiki

at system resource PCI it detect the card..

but the wireless interface still not showing.

aliaslaci · Thu May 28, 2009 11:29 pm

Hi! (excuse me for my english)
Can someone do a test where the SNR is little more ...how to say it... realistic

(12-20-25dBm)?

tyronzn · Fri May 29, 2009 10:59 am

hello

yes i used two antennas on each card,i did an outdoor test over a couple hundred meters using a dual polarized antenna with RB433AH's either side using the SR71A's.I also had two pc's connected to each board on both ends and used the mikrotik bandwidth test to do the tests through the boards for accurate results.I Suspect i might have been able to get faster speeds as the 433AH's only have 10/100 ethernet ports as compared to the RB600A's with Gigabit.Despite that i managed to consistently get 95.7Mbps without any drops.In my opinion i think the new R52N would perform a bit better though
Tyronzn
95.7Mbps UDP ? TCP ? packet size ?

It was a UDP test one way "receive" with a packet size of 1500 bytes

fly · Fri May 29, 2009 2:51 pm

Nstreme enabled!!!

fly · Fri May 29, 2009 2:59 pm

I know Nstreme should not work, but I can see big differences with and without.
More bandwidth, more stability... just like a 802.11a link.

2jarek · Fri May 29, 2009 3:05 pm

But I can see big differences with and without.
More bandwidth, more stability... just like a 802.11a link.

RB411AH has 100mbit fdx ethernet..... how possible >180 mbit/s in UDP ????

magnavox · Fri May 29, 2009 3:06 pm

Nstreme enabled!!!

Big!

Are you running a UBNT card... SR71-A?
Link distance?
Number of antennas: x2 or x3?

fly · Fri May 29, 2009 3:08 pm

Nstreme enabled!!!
Big!

Are you running a UBNT card... SR71-A?
Link distance?
Number of antennas: x2 or x3?

Yes, UBNT SR71-A, I'm waiting for R52N...
2 antennas, 100m.

triztrz · Fri May 29, 2009 3:21 pm

something weird...

the 2 card i purchase yesterday.. still dont show at wireless interface..
and today i go to purchase another one with.. same series TL-WN851N .. and it's detected....

Fri May 29, 2009 3:24 pm

if not detected, email support the supout.rif file

und3ath · Fri May 29, 2009 4:09 pm

But I can see big differences with and without.
More bandwidth, more stability... just like a 802.11a link.
RB411AH has 100mbit fdx ethernet..... how possible >180 mbit/s in UDP ????

because he is testing wireless to wireless

Fri May 29, 2009 4:10 pm

But I can see big differences with and without.
More bandwidth, more stability... just like a 802.11a link.
RB411AH has 100mbit fdx ethernet..... how possible >180 mbit/s in UDP ????

He does not test Ethernet.

Stefan

tyronzn · Fri May 29, 2009 4:37 pm

Nstreme enabled!!!
Big!

Are you running a UBNT card... SR71-A?
Link distance?
Number of antennas: x2 or x3?

+- 700 meters using the UBNT SR71A cards.I used two antennas either side,to be more technical i was using a dual polarized panel antenna meaning it had two connectors on it which in turn had to pig tails plugged into either card.I've also
tested the Compex WLM200NX in RouterOS which works.

Fri May 29, 2009 5:52 pm

The current nstreme implementation isn't so stable that's why we are not recomending to use it. As soon as we will fix it we will announce it.

Gerard · Fri May 29, 2009 6:09 pm

yay!

We just got our shipment R52N's and R2N's.. I'm about to try 2 mile link with dual pol arc panels. I'll post back my results..

Gerard

frontiersteve · Fri May 29, 2009 7:15 pm

This may not be the place to make a board suggestion, but it would be nice to have a gigE port option on 400 series routers for the 802.11n that is now supported.

mp3turbo · Fri May 29, 2009 9:27 pm

Gerard, together with your 802.11n result please post maximum speeds and distances you were able to reach with standard 802.11a. This will give us little bit overview about your wireless skills, not to sound offending. I mean this seriously, in a good way.

Because there are many people here on this board who are unable to pass 20Mbit+ wireless throughput. And I don't consider any new-technology results from them to be of any importance. You know what I mean... Once again, I don't want to sound rude or rough...

edit : you are certified Mikrotik guy and I expect nothing just excellent results. My hopes are really really high, we need fast and I mean really fast wireless technology like a salt. Technical advancements of 802.11n, specifically speaking more subcarriers, forward error correction 5/6 instead of 3/4, lower MAC overhead and shorter guard interval should bring approx. 50% more real performance compared to 802.11a. I'm eagerly waiting for any real-world results and I'm not talking about "in my home, I achieved...", that's not interesting for me. Your 2 mile shot will be the first posted here I guess...

omega-00 · Sat May 30, 2009 7:22 pm

But I can see big differences with and without.
More bandwidth, more stability... just like a 802.11a link.
RB411AH has 100mbit fdx ethernet..... how possible >180 mbit/s in UDP ????

He's doing the test between the units themselves, not between computers on each end.

angboontiong · Sun May 31, 2009 10:40 am

possible to post the test result with only tcp traffic as well the ping time during the testing period.

Mon Jun 01, 2009 8:54 am

This may not be the place to make a board suggestion, but it would be nice to have a gigE port option on 400 series routers for the 802.11n that is now supported.

that's why we make the RB600

xezen · Mon Jun 01, 2009 11:16 am

but the rb600 is a waste if you have 3 pc on it
thats why the request is for the rb400

to much power for what it is used for?

Mon Jun 01, 2009 11:17 am

so the only problem with RB600 is that it has 4 miniPCI slots insteads of 3 ?

we can't make a specific device for each customer needs, it's very hard to make everyone happy.

xezen · Mon Jun 01, 2009 11:21 am

thats true!

xezen · Mon Jun 01, 2009 11:22 am

does anyone have information on the antennas that were used on the tests?

so we can look at the hardwer side of things?

und3ath · Mon Jun 01, 2009 1:30 pm

so the only problem with RB600 is that it has 4 miniPCI slots insteads of 3 ?

we can't make a specific device for each customer needs, it's very hard to make everyone happy.

I think he thought, that it is a waste of RB600 power because of 3 LAN connected PCs

danletkeman · Tue Jun 02, 2009 4:35 am

An RB411AH "G" would be nice because we would not need to replace all of our existing enclosures. We are using them mostly for backhaul's and only need one card anyway.

Qusay · Tue Jun 02, 2009 11:06 am

RB433AH-Gigae thernet

I agree .....

complete2006 · Tue Jun 02, 2009 2:05 pm

Yes. Same board size but with GBit. RB600 is to big to fit in the Mars antennas.

karyal · Tue Jun 02, 2009 5:10 pm

This may not be the place to make a board suggestion, but it would be nice to have a gigE port option on 400 series routers for the 802.11n that is now supported.
that's why we make the RB600

Normis, isn't the RB433AH supposed to have a faster cpu then the RB600? generally speaking my experience is we can achieve much better throughput using 433AH than RB600 (i got no more than 52/56 mpbs TCP on RB600, but up to 72 with RV433AH), on exactly the same configuration, and i guess this is related to a faster cpu.. I'm afraid that with more traffic handled by the radio the RB cpu will be the bottleneck.
And of course the board size of a 433AH is quite an advantage over RB600, not to mention it's quite cheaper.
I would love an RB443AH with gigabit ethernet too, and i agree with the collegue who asked MT to think about it.
A quick possible solution for us with 433AH may be "simply" LAG bonding the two (or three) FE interfaces on the board.

mp3turbo · Tue Jun 02, 2009 5:12 pm

gentlemen, could we please DEVOTE this topic to 802.11n achievements? General talk about gigabit interfaces and sizes of board does not belong here.

I will kindly ask anybody with real-world 802.11n results to post their performance. I'm eagerly waiting delivery of my cards which has been postponed AGAIN

karyal · Tue Jun 02, 2009 5:34 pm

gentlemen, could we please DEVOTE this topic to 802.11n achievements? General talk about gigabit interfaces and sizes of board does not belong here.

I agree with you on the "size", but not on the interfaces part.
I think that having 802.11n being able to run at 200/300Mbps on the air is great, but actually being able to use that 200/300mbps is part of the "achievement" we are all asking to 802.11n, as btw having 802.11n being able to run stable with nstreme (if 802.11n behaves the same as 802.11a) is quite vital on long distance links

you can hardly run a 20km link at full speed with nstreme turned off (while it quite surely would work at 100mt)

Strictly speaking though, my question is: is the RB600 cpu fast enough to take full advantage of 802.11n or do we need something faster like an RB433AH?

Thanks,
Ricky

kapulan · Tue Jun 02, 2009 7:24 pm

Supported the SR71-A card the beta3? I test this card demo beta3 and not show the interface list.

mp3turbo · Tue Jun 02, 2009 7:49 pm

karyal, my remarks:

- forget about 300Mbps in air, we will consider ourselves unbelievably lucky with ~130Mbps on 40MHz channel

- I fully agree with you that you have to supply enough data for such speeds : you know for sure that you need gigabit interface. Still I believe that whether CPU of routerboard XY is strong enough or not does not belong to this topic, feel free to open your own topic and discuss cpu strength, connection tracking, routing/bridging and other performance topics there.

I don't want to sound rude, but this discussion is getting definitely out of "802.11n" name it has in subject. In a moment, we will be discussing whether ConnTrack should be off and what is the performance hit, number of packets that can go thru routerboard AB and XY and similar chapters that have NOTHING IN COMMON with 802.11n.

Please keep this topic PURELY 802.11n related. You know for sure you need a lot of power, so buy dualcore x86 platform and it will be sufficient to supply data needed for 802.11n. Nobody knows for sure now if those small underpowered soap-boxes can supply it and no one tried it in real world already. You can't use x86? So wait please until somebody performs thorough test or do it yourself and post results in separate thread.

802.11n only here, please. No routerboard discussion. No packets per second, no power supply issues... etc. Let's have at least one topic FOCUSED.

magnavox · Tue Jun 02, 2009 8:44 pm

Supported the SR71-A card the beta3? I test this card demo beta3 and not show the interface list.

If you use Routerboard, update all: licence key and Routerboard BIOS.

karyal · Tue Jun 02, 2009 8:53 pm

mp3turbo, i don't want to sound rude either, i agree with you that is a good thing to stay on-topic, but please note this is a RouterOS Wireless Networking Section, on Mikrotik forum, where some started the topic mentioning
"I would be great if the first people to test the new 802.11n support in RouterOS v4 beta 3 would post their experiences here in this thread."

It sounds to me a pretty clear message of "use 802.11n on Routeros, and hardware supporting it, and test how it is".
No strict 802.11n design or theory, this is no IEEE forum, is Mikrotik.. is about how fast is 802.11n running on Mikrotik, is how fast can 802.11n run on Routerboard or MT supported hardware.. i guess that most of the people (here) don't really care too much if 802.11n runs unbelievably fast on any platform that is not MT and ROS related, and if i want to do my part in testing "the new 802.11n support in RouterOS v4 beta 3" and "post my experience here", i would like to do that by knowing, once i get my hands on the new MT radios, which RB is better to use (or not to use), and report back to MT

mp3turbo · Tue Jun 02, 2009 9:38 pm

ok, I give up. I agree with you: "It sounds to me a pretty clear message of "use 802.11n on Routeros, and hardware supporting it, and test how it is"...

so it should look like:

- I achieved 120Mbps, 2x mikrotik r52n cards, 4 miles distance, 2x 26dB antennas, -70dB rx/tx reported signal, Canada Ontario rural environment, RB433AH on one side, RB600 on second side, 5480MHz, MCS7, testing wireless to wireless Tcp only (no real data, no computers available behind both routers)

or something like that. This topic should not look like how it looks now:

- "please please can we have gigabit interface on RB433xxx"...
- "we need small gigabit rb600xxx"
- "it's not possible to put XXXX into antenna YYYY"

such messages have NOTHING in common with 802.11n.

ihernandez · Wed Jun 03, 2009 5:36 pm

Hello, I would like to know if I can use this cards with external dish antennas for point to point links

Wed Jun 03, 2009 5:57 pm

Hello, I would like to know if I can use this cards with external dish antennas for point to point links

Why not? As I've learned max separation should be achieved. So use one H and one V-Pol and
separate them.

Stefan

infowest · Wed Jun 03, 2009 8:23 pm

Has anyone tried this with a dual-polarity dish? Any real benefit/cost doing it that way vs using two separate antennas?

Randy

jmpiorno · Wed Jun 03, 2009 10:05 pm

Hello friends
Someone has tried 802.11n for multi-point links? without line of sight? that results obtained?

phendry · Thu Jun 04, 2009 2:52 pm

Has anyone tried this with a dual-polarity dish? Any real benefit/cost doing it that way vs using two separate antennas?

Randy

Hoping to on a 10km link next week with Dual-pol MTi but still waiting for the R52N's too arrive.

jsuter · Thu Jun 04, 2009 9:53 pm

such messages have NOTHING in common with 802.11n.

You've had absolutely nothing to add to this thread concerning your own performance results. Where are YOUR benchmarks?

If you want to control how the content is delivered to you I suggest hiring a consulting firm to type you up a report on how Mikrotik's 11n works for them. It'll actually give you the right to dictate how the information is handed to you.

Otherwise - feel free quit whining about the FREE information you're receiving from this forum. If the post doesn't have info you want, simply skip over it. When there's actually *real* information worth posting it'll get posted in the WIKI.

xezen · Fri Jun 05, 2009 7:51 am

such messages have NOTHING in common with 802.11n.
You've had absolutely nothing to add to this thread concerning your own performance results. Where are YOUR benchmarks?

If you want to control how the content is delivered to you I suggest hiring a consulting firm to type you up a report on how Mikrotik's 11n works for them. It'll actually give you the right to dictate how the information is handed to you.

Otherwise - feel free quit whining about the FREE information you're receiving from this forum. If the post doesn't have info you want, simply skip over it. When there's actually *real* information worth posting it'll get posted in the WIKI.

thats true but many people cant just get the stuff in two min
example tplink n card has been out for some time and im still waiting

n has just come out from mikrotik and i must still wait 4 week to get one so someone like me
cant put down information on this thats why becouse i have nothing to play with

thats why outher topics come up
becouse we cant talk about the real topic but we can give ideas to improve the topic with the information we have read on the form and the wiki
becouse thats the point to the form right
help others improve on it (MK Team)
and others fix problems

karyal · Fri Jun 05, 2009 7:43 pm

Just spent some hours testing the new radios (r52n)

Single radio setup, dual antennas, cross polarized, rb433AH, 5ghz, nstream enabled

Tried to show as many details as possible.

Udp tx test from board to board

testn2.JPG

UDP rx test, board to board

test-n3b.jpg

UDP full duplex test

test-n4b.jpg

karyal · Fri Jun 05, 2009 7:56 pm

full duplex UDP traffic generated from external boards

test-n5-100mbit-offboard.JPG

A couple of notes... tried to load balance the traffic, on external boards, by bonding two ethernet and avoiding the single ethernet 100mbps limit of FE.
Traffic gets balanced equally, but still tops at 100mbps total.. don't know if it's a bug or simply me not being able to configure bonding right on MT/MT setup. I usually configured LAG by port aggregation on the switch + device, this time i had no switch, so u isued bonding rr.

Nstream seems to be stable, but TCP traffic performance is not exciting as UDP ones.
Topped at 78/80 Mbps HDX no matter of the boards, way we measured and setup..
Seems no real gain compared to traditional 802.11a turbo
Will test at longer distance (15Km) next week.

infowest · Fri Jun 05, 2009 8:21 pm

Thanks for the tests. What were your HT settings?

karyal · Fri Jun 05, 2009 8:25 pm

Thanks for the tests. What were your HT settings?

AMPDU: All selected
HT Extension Channel: Above Control
HT TX/RX Chains: 0 1

meshnet · Fri Jun 05, 2009 9:01 pm

Were these at any kind of distance?
Tested indoors, outdoors?

Thanks,
Richard

karyal · Fri Jun 05, 2009 9:09 pm

Were these at any kind of distance?
Tested indoors, outdoors?

Thanks,
Richard

We tested at short distance, tried with both panels and mini-omni. same results.
I was quite interested to see the max speed that could be reached in an ideal environment to move later on a bigger distance and see the impact, and was interested in checking how stable was nstream + n.
It seems to be stable indeed, we had the boards running for all the afternoon without a single problem or link drop.
I feel that much of the advantage on speed is because of signal rebuild that is done.
I don't know how much it can be seen by the screenshots, but CPU power on the receiving side is still quite high, much higher than plain 5ghz... guess it may be because of the cpu working on signals.

kapulan · Fri Jun 05, 2009 9:39 pm

full duplex UDP traffic generated from external boards

test-n5-100mbit-offboard.JPG
A couple of notes... tried to load balance the traffic, on external boards, by bonding two ethernet and avoiding the single ethernet 100mbps limit of FE.
Traffic gets balanced equally, but still tops at 100mbps total.. don't know if it's a bug or simply me not being able to configure bonding right on MT/MT setup. I usually configured LAG by port aggregation on the switch + device, this time i had no switch, so u isued bonding rr.

Nstream seems to be stable, but TCP traffic performance is not exciting as UDP ones.
Topped at 78/80 Mbps HDX no matter of the boards, way we measured and setup..
Seems no real gain compared to traditional 802.11a turbo
Will test at longer distance (15Km) next week.

Nice speed with udp.You can running the test with tcp packet?Please paste the resoult.
Mt support: The "n" mode support x86 system with minipci to pci card,or only routerboard platform?

karyal · Fri Jun 05, 2009 9:52 pm

Nice speed with udp.You can running the test with tcp packet?Please paste the resoult.

I already posted the results with TCP

Nstream seems to be stable, but TCP traffic performance is not exciting as UDP ones.
Topped at 78/80 Mbps HDX no matter of the boards, way we measured and setup..

flavio · Sat Jun 06, 2009 6:15 pm

Hi,
If I have only one 5 Ghz channel available e one antenna each side, can I make a 802.11n link (8 km)? Expected performance of this link? Thanks.

roc-noc.com · Sun Jun 07, 2009 1:31 am

My Mikrotik "n" cards are arriving on Monday so I have not actually tested this yet. But I believe you will not see any improvement over 802.11a with this link. You need at least two antennas at each end for mimo (n).

Tom

Casket · Sun Jun 07, 2009 3:20 am

Advantages and disadvantages of dual polarized vs. two separate antennas each end? in ptp.

jsuter · Mon Jun 08, 2009 8:39 am

My Mikrotik "n" cards are arriving on Monday so I have not actually tested this yet. But I believe you will not see any improvement over 802.11a with this link. You need at least two antennas at each end for mimo (n).

Tom

MIMO exists to lower effective EVM and other distortions. Technically MIMO is (now, at least in the newer drafts) part of the pre-standard but its not actually required for an 11n link (at least at HT20 rates). It sounds like (BUT I HAVE NO PERSONAL EXPERIENCE YET) it requires 2 antennas to run HT40 rates, but that's a limitation of Atheros's chipset, not the protocol itself.

I'll be ordering some 11n cards from you in the next few days. I want to test the 11G MIMO modes for some APs. I've also read the AR71xx platform (the RB400-series CPU) was designed for Atheros's 11N cards, and actually requires kernel hacks to work around bus errors to support the AR5xxx series radio chipsets. If this is true, I suspect switching my busy+crashy RB433H sites over ASAP.

xezen · Mon Jun 08, 2009 3:56 pm

can you please send more information on the anttenna that you used were thay boght or home made?

karyal · Tue Jun 09, 2009 12:26 am

can you please send more information on the anttenna that you used were thay boght or home made?

Bought (Elboxrf)

hazardz · Tue Jun 09, 2009 6:41 am

just passing by, its verry insteresting..
i was waiting for this kind of moment for Mikortik and The 802.11n compability..

sorry could not test the device becouse of my low budgets..
i'm verry excited and looking forward to for the result...

Tue Jun 09, 2009 12:07 pm

full duplex UDP traffic generated from external boards

test-n5-100mbit-offboard.JPG
A couple of notes... tried to load balance the traffic, on external boards, by bonding two ethernet and avoiding the single ethernet 100mbps limit of FE.
Traffic gets balanced equally, but still tops at 100mbps total.. don't know if it's a bug or simply me not being able to configure bonding right on MT/MT setup. I usually configured LAG by port aggregation on the switch + device, this time i had no switch, so u isued bonding rr.

Nstream seems to be stable, but TCP traffic performance is not exciting as UDP ones.
Topped at 78/80 Mbps HDX no matter of the boards, way we measured and setup..
Seems no real gain compared to traditional 802.11a turbo
Will test at longer distance (15Km) next week.

Got same results without nstreme. But in my tests Nstreme was not stable.
I got reconnects. Without nstreme I've stable connection but performance
is not as stable as yours. May be I`ve some interference.
Tested with 2 Stella Doradus Dual Poles in lab (reduced tx-power)
RB411AH/R52N.

Stefan

andis · Tue Jun 09, 2009 4:08 pm

indoor 802.11n test (~10m range, wooden wall obstruction) done with R52n cards, RB/493AH boards and original indoor ac/swi antennas.

both RouterBoards running RouterOS v4.0beta3.

bandwidth test results between both boards via radio are the following:

UDP simple direction 223.5 Mbps
UDP full duplex 102.8 + 102.3 = 205.1 Mbps

### AccessPoint config ###

/interface wireless
set 0 ack-timeout=dynamic adaptive-noise-immunity=none allow-sharedkey=no antenna-gain=0 \
antenna-mode=ant-a area="" arp=enabled band=5ghz-onlyn basic-rates-a/g=6Mbps basic-rates-b=\
1Mbps comment="" compression=no country=no_country_set default-ap-tx-limit=0 \
default-authentication=yes default-client-tx-limit=0 default-forwarding=yes dfs-mode=none \
disable-running-check=no disabled=no disconnect-timeout=3s frame-lifetime=0 frequency=5180 \
frequency-mode=manual-txpower hide-ssid=no ht-ampdu-priorities=0,1,2,3,4,5,6,7 ht-amsdu-limit=\
8192 ht-amsdu-threshold=8192 ht-basic-mcs=mcs-0 ht-extension-channel=above-control \
ht-guard-interval=any ht-rxchains=0,1 ht-supported-mcs="mcs-0,mcs-1,mcs-2,mcs-3,mcs-4,mcs-5,mcs\
-6,mcs-7,mcs-8,mcs-9,mcs-10,mcs-11,mcs-12,mcs-13,mcs-14,mcs-15" ht-txchains=0,1 \
hw-fragmentation-threshold=disabled hw-protection-mode=none hw-protection-threshold=0 \
hw-retries=4 mac-address=00:1D:0F:BB:8C:66 max-station-count=2007 mode=ap-bridge mtu=1500 \
name=wlan1 on-fail-retry-time=100ms periodic-calibration=default \
periodic-calibration-interval=60 preamble-mode=both proprietary-extensions=post-2.9.25 \
radio-name=001D0FBB8C66 rate-set=configured scan-list=default security-profile=default ssid=\
router-lv station-bridge-clone-mac=00:00:00:00:00:00 supported-rates-a/g=6Mbps,54Mbps \
supported-rates-b=1Mbps,2Mbps,5.5Mbps,11Mbps tx-power-mode=default update-stats-interval=\
disabled wds-cost-range=50-150 wds-default-bridge=none wds-default-cost=100 wds-ignore-ssid=no \
wds-mode=disabled wmm-support=enabled
/interface wireless manual-tx-power-table
set wlan1 comment="" manual-tx-powers="1Mbps:17,2Mbps:17,5.5Mbps:17,11Mbps:17,6Mbps:17,9Mbps:17,12M\
bps:17,18Mbps:17,24Mbps:17,36Mbps:17,48Mbps:17,54Mbps:17,HT20-1:0,HT20-2:0,HT20-3:0,HT20-4:0,HT\
20-5:0,HT20-6:0,HT20-7:0,HT20-8:0,HT40-1:0,HT40-2:0,HT40-3:0,HT40-4:0,HT40-5:0,HT40-6:0,HT40-7:\
0,HT40-8:0"
/interface wireless nstreme
set wlan1 comment="" disable-csma=no enable-nstreme=no enable-polling=yes framer-limit=3200 \
framer-policy=none
/interface wireless align
set active-mode=yes audio-max=-20 audio-min=-100 audio-monitor=00:00:00:00:00:00 filter-mac=\
00:00:00:00:00:00 frame-size=300 frames-per-second=25 receive-all=no ssid-all=no
/interface wireless sniffer
set channel-time=200ms file-limit=10 file-name="" memory-limit=10 multiple-channels=no \
only-headers=no receive-errors=no streaming-enabled=no streaming-max-rate=0 streaming-server=\
0.0.0.0
/interface wireless snooper
set channel-time=200ms multiple-channels=yes receive-errors=no

### Station config ###

/interface wireless
set 0 ack-timeout=dynamic adaptive-noise-immunity=none allow-sharedkey=no antenna-gain=0 antenna-mode=ant-a area="" arp=\
enabled band=5ghz-onlyn basic-rates-a/g=6Mbps basic-rates-b=1Mbps comment="" compression=no country=no_country_set \
default-ap-tx-limit=0 default-authentication=yes default-client-tx-limit=0 default-forwarding=yes dfs-mode=none \
disable-running-check=no disabled=no disconnect-timeout=3s frame-lifetime=0 frequency=5180 frequency-mode=\
manual-txpower hide-ssid=no ht-ampdu-priorities=0,1,2,3,4,5,6,7 ht-amsdu-limit=8192 ht-amsdu-threshold=8192 \
ht-basic-mcs=mcs-0 ht-extension-channel=above-control ht-guard-interval=any ht-rxchains=0,1 ht-supported-mcs=\
mcs-0,mcs-1,mcs-2,mcs-3,mcs-4,mcs-5,mcs-6,mcs-7,mcs-8,mcs-9,mcs-10,mcs-11,mcs-12,mcs-13,mcs-14,mcs-15 ht-txchains=0,1 \
hw-fragmentation-threshold=disabled hw-protection-mode=none hw-protection-threshold=0 hw-retries=4 mac-address=\
00:1D:0F:BB:8C:1E max-station-count=2007 mode=station mtu=1500 name=wlan1 on-fail-retry-time=100ms \
periodic-calibration=default periodic-calibration-interval=60 preamble-mode=both proprietary-extensions=post-2.9.25 \
radio-name=001D0FBB8C1E rate-set=configured scan-list=default security-profile=default ssid=router-lv \
station-bridge-clone-mac=00:00:00:00:00:00 supported-rates-a/g=6Mbps,54Mbps supported-rates-b=\
1Mbps,2Mbps,5.5Mbps,11Mbps tx-power-mode=default update-stats-interval=disabled wds-cost-range=50-150 \
wds-default-bridge=none wds-default-cost=100 wds-ignore-ssid=no wds-mode=disabled wmm-support=enabled
/interface wireless manual-tx-power-table
set wlan1 comment="" manual-tx-powers="1Mbps:17,2Mbps:17,5.5Mbps:17,11Mbps:17,6Mbps:17,9Mbps:17,12Mbps:17,18Mbps:17,24Mbps:1\
7,36Mbps:17,48Mbps:17,54Mbps:17,HT20-1:0,HT20-2:0,HT20-3:0,HT20-4:0,HT20-5:0,HT20-6:0,HT20-7:0,HT20-8:0,HT40-1:0,HT40-2:\
0,HT40-3:0,HT40-4:0,HT40-5:0,HT40-6:0,HT40-7:0,HT40-8:0"
/interface wireless nstreme
set wlan1 comment="" disable-csma=no enable-nstreme=no enable-polling=yes framer-limit=3200 framer-policy=none
/interface wireless align
set active-mode=yes audio-max=-20 audio-min=-100 audio-monitor=00:00:00:00:00:00 filter-mac=00:00:00:00:00:00 frame-size=\
300 frames-per-second=25 receive-all=no ssid-all=no
/interface wireless sniffer
set channel-time=200ms file-limit=10 file-name="" memory-limit=10 multiple-channels=no only-headers=no receive-errors=no \
streaming-enabled=no streaming-max-rate=0 streaming-server=0.0.0.0
/interface wireless snooper
set channel-time=200ms multiple-channels=yes receive-errors=no

jsuter · Wed Jun 10, 2009 6:19 am

Is there any plan to backport 11n support into RouterOS 3.xx?

Thanks!

tritsako · Wed Jun 10, 2009 6:48 am

Does anybody tested RB433 (and OS Ver. 3.10) with R52n and 802.11n connection?

Thanks
Costas

Wed Jun 10, 2009 8:44 am

Does anybody tested RB433 (and OS Ver. 3.10) with R52n and 802.11n connection?

no, because 802.11n works only on v4

jsuter · Wed Jun 10, 2009 8:48 am

Does anybody tested RB433 (and OS Ver. 3.10) with R52n and 802.11n connection?
no, because 802.11n works only on v4

Is there any plan to add 11n support to v3 in the future?

Wed Jun 10, 2009 8:53 am

when is a new version a "new version" if all it's features are backported? the old one becomes the new one, and the new one becomes identical. we could just remove any version numbers, and just tell you when features are introduced. I don't think it's a good idea. Version numbers are there for a reason

jsuter · Wed Jun 10, 2009 9:40 am

when is a new version a "new version" if all it's features are backported? the old one becomes the new one, and the new one becomes identical. we could just remove any version numbers, and just tell you when features are introduced. I don't think it's a good idea. Version numbers are there for a reason

So what you're saying is "if you've got routerboards not licensed to upgrade to 4.x you're not going to be running 11n on them without shilling out more than the boards themselves are worth for a 4.x license?" - am I getting this correctly?

Many of my 4xx's won't upgrade past 3.x. Some of these boards are less than a year old. Wheres MT's upgrade license policy?

Wed Jun 10, 2009 9:48 am

this should not be true, all Level4 licenses allow upgrade to v4.xx if they were purchased when v3 was already released.

if not, the upgraded key will fix it (when you upgrade to v4beta, it makes a new key, and makes sure that your upgrade times are correct).

jsuter · Wed Jun 10, 2009 10:24 am

this should not be true, all Level4 licenses allow upgrade to v4.xx if they were purchased when v3 was already released.

if not, the upgraded key will fix it (when you upgrade to v4beta, it makes a new key, and makes sure that your upgrade times are correct).

What about L3 that only show to upgrade to 3.x? What are the options on these boards?

Wed Jun 10, 2009 10:25 am

As I said, this shouldn't be so, and the key upgrade in v4beta will fix it for you

conjurer · Wed Jun 10, 2009 10:59 am

Can anyone tell me, what type of channel 5ghz onlyN uses? Lets say i set it to work in 5180, than it covers 5170-5190 or 5160-5200? By specification 300 Mbps is with 2 antennas, both turbo channel. Will channel width be ajustable in future?

Wed Jun 10, 2009 11:46 am

Can anyone tell me, what type of channel 5ghz onlyN uses? Lets say i set it to work in 5180, than it covers 5170-5190 or 5160-5200? By specification 300 Mbps is with 2 antennas, both turbo channel. Will channel width be ajustable in future?

The channels you configure are the middle frequency of the 20MHz Channel. So 5180 is 5170-5190.
When you configure the second channel above you use 5170-5210. When you do not enable the
second channel on both devices you stay at 20MHz Channel.

Attached a wi-spy graphic running bandwith-test 5500MHz second channel above.

Stefan

Wed Jun 10, 2009 11:49 am

that's why it's important to make sure your nearby AP frequencies are not overlapping.

examples attached

tritsako · Wed Jun 10, 2009 4:53 pm

Thank you.

manners · Wed Jun 10, 2009 11:07 pm

Hi everyone

i currently have 2 of the r52n cards on rb411ah boards and seem to have hit a barrier at 60mbps udp throughput,

which to me means im missing something i have completely followed the example setup and still end up with the same results, and i have noticed by some of the pics that people have there cards connected at 216mbps

i would love to know how this is possible as mine never go above 65mbps ht

any help appreciated

conjurer · Thu Jun 11, 2009 2:31 am

The channels you configure are the middle frequency of the 20MHz Channel. So 5180 is 5170-5190.
When you configure the second channel above you use 5170-5210. When you do not enable the
second channel on both devices you stay at 20MHz Channel.

So above control and below control settings do add 20 mhz channel.

Thu Jun 11, 2009 10:10 am

The channels you configure are the middle frequency of the 20MHz Channel. So 5180 is 5170-5190.
When you configure the second channel above you use 5170-5210. When you do not enable the
second channel on both devices you stay at 20MHz Channel.
So above control and below control settings do add 20 mhz channel.

yes

Thu Jun 11, 2009 10:13 am

the next RouterOS release will have additional information about transmit power, it will make things more clear

yes, what you wrote is right

for two chains, this is what it will look like (real tx power is different for XR cards):

tx-power-n.PNG

rpress · Fri Jun 12, 2009 8:26 pm

Should 4.0beta3 work with AR5416 cards? I tried mine and it does not create an interface. It is shown in PCI resources:

name: AR5416 802.11a/b/g/n Wireless PCI Adapter (rev: 1)
vid: 0x168c
did: 0x168c

Houge · Fri Jun 12, 2009 9:00 pm

We would like to share some of our findings using the new R52N cards.

The setup is RB300 and RB600 with ROS4beta3, 1 dishantenna at each point.

For about 2 years now we have used R52H cards, and gotten a good speed ~25mbit TCP on a 4km link. (2.4GHz turbomode)
Last week we got 2x R52N cards, and hoped for a better link and speeds. But its very dissapointing on 2.4GHz B/G/N performance .
On R52H we used turbomode (40mhz), but the new ones don´t seem to be able to work at 40MHz on 1 rx/tx chain.

Why is it not possible to use 40MHz channels on R52N, like you can on R52H?
Its locked at max 65Mbit when using 1 tx/rx chain, and we tryed alot of settings. (local lab test)
The HT extension channel does absolutely nothing when you use 1 rx/tx chain. Is it a bug?

And could someone explain what the difference between 2.4Ghz b/g/n and 2.4Ghz-only-N is, as I see absolutely no difference, it still uses the same B rates when in only-N mode.

Another odd thing is that it seems it only uses long guard interval 800ns, since the rate is at 65Mbit, and not 72.2Mbit.
How do you change it to short 400ns? You can only change from any to long, and that does not change anything.

With the N card in 2.4GHz B/G/N mode we are now only getting MAX 10Mbit on our link, compaired to ~25Mbit in 2.4Ghz turbomode on R52H. At this point, it seems to be a total waste of money.

And 10Mbit is only if you force it to a higher transfer rate. If the data rate is in default, we are getting no more than 5Mbit. It seems like the ROS dont higher the transfer rate, it keeps switching from 1,2,5.5,11mbit. We got about 30dB SNR on the link.

The only positive thing is a 5dB higher signal strenth and a 10-12dB lower noice floor, which makes us wonder even more, why the transfer speed are so bad, compaired to our old R52H in 2,4Ghz turbomode. The total SNR is 5-10dB better then our R52H, but speeds are 15Mbit lower. That dont make any sense.

Does someone have any answers to this, or some data they made on only 1 rx/tx chain?

Regards
Brian.

roc-noc.com · Sat Jun 13, 2009 12:23 am

We would like to share some of our findings using the new R52N cards.

The setup is RB300 and RB600 with ROS4beta3, 1 dishantenna at each point.

For about 2 years now we have used R52H cards, and gotten a good speed ~25mbit TCP on a 4km link. (2.4GHz turbomode)
Last week we got 2x R52N cards, and hoped for a better link and speeds. But its very dissapointing on 2.4GHz B/G/N performance .
On R52H we used turbomode (40mhz), but the new ones don´t seem to be able to work at 40MHz on 1 rx/tx chain.

Why is it not possible to use 40MHz channels on R52N, like you can on R52H?
Its locked at max 65Mbit when using 1 tx/rx chain, and we tryed alot of settings. (local lab test)
The HT extension channel does absolutely nothing when you use 1 rx/tx chain. Is it a bug?

And could someone explain what the difference between 2.4Ghz b/g/n and 2.4Ghz-only-N is, as I see absolutely no difference, it still uses the same B rates when in only-N mode.

Another odd thing is that it seems it only uses long guard interval 800ns, since the rate is at 65Mbit, and not 72.2Mbit.
How do you change it to short 400ns? You can only change from any to long, and that does not change anything.

With the N card in 2.4GHz B/G/N mode we are now only getting MAX 10Mbit on our link, compaired to ~25Mbit in 2.4Ghz turbomode on R52H. At this point, it seems to be a total waste of money.

And 10Mbit is only if you force it to a higher transfer rate. If the data rate is in default, we are getting no more than 5Mbit. It seems like the ROS dont higher the transfer rate, it keeps switching from 1,2,5.5,11mbit. We got about 30dB SNR on the link.

The only positive thing is a 5dB higher signal strenth and a 10-12dB lower noice floor, which makes us wonder even more, why the transfer speed are so bad, compaired to our old R52H in 2,4Ghz turbomode. The total SNR is 5-10dB better then our R52H, but speeds are 15Mbit lower. That dont make any sense.

Does someone have any answers to this, or some data they made on only 1 rx/tx chain?

Regards
Brian.

Normis answered that question here: http://forum.mikrotik.com/viewtopic.php?f=7&t=32549

The R52n does not support turbo mode. I don't think that is in the 802.11n draft specifications.
I don't think it was in 802.11a/g either but I seem to remember that atheros added Turbo on their own outside of the standards.

Tom

RK · Sat Jun 13, 2009 2:03 am

Does someone have any answers to this, or some data they made on only 1 rx/tx chain?

This was answered before. You need to connect TWO antennas to the N radio card, not just one.
If you only have one antenna on each side, use the R5H in turbo mode instead.

Houge · Sat Jun 13, 2009 1:13 pm

We would like to share some of our findings using the new R52N cards.

The setup is RB300 and RB600 with ROS4beta3, 1 dishantenna at each point.

For about 2 years now we have used R52H cards, and gotten a good speed ~25mbit TCP on a 4km link. (2.4GHz turbomode)
Last week we got 2x R52N cards, and hoped for a better link and speeds. But its very dissapointing on 2.4GHz B/G/N performance .
On R52H we used turbomode (40mhz), but the new ones don´t seem to be able to work at 40MHz on 1 rx/tx chain.

Why is it not possible to use 40MHz channels on R52N, like you can on R52H?
Its locked at max 65Mbit when using 1 tx/rx chain, and we tryed alot of settings. (local lab test)
The HT extension channel does absolutely nothing when you use 1 rx/tx chain. Is it a bug?

And could someone explain what the difference between 2.4Ghz b/g/n and 2.4Ghz-only-N is, as I see absolutely no difference, it still uses the same B rates when in only-N mode.

Another odd thing is that it seems it only uses long guard interval 800ns, since the rate is at 65Mbit, and not 72.2Mbit.
How do you change it to short 400ns? You can only change from any to long, and that does not change anything.

With the N card in 2.4GHz B/G/N mode we are now only getting MAX 10Mbit on our link, compaired to ~25Mbit in 2.4Ghz turbomode on R52H. At this point, it seems to be a total waste of money.

And 10Mbit is only if you force it to a higher transfer rate. If the data rate is in default, we are getting no more than 5Mbit. It seems like the ROS dont higher the transfer rate, it keeps switching from 1,2,5.5,11mbit. We got about 30dB SNR on the link.

The only positive thing is a 5dB higher signal strenth and a 10-12dB lower noice floor, which makes us wonder even more, why the transfer speed are so bad, compaired to our old R52H in 2,4Ghz turbomode. The total SNR is 5-10dB better then our R52H, but speeds are 15Mbit lower. That dont make any sense.

Does someone have any answers to this, or some data they made on only 1 rx/tx chain?

Regards
Brian.
Normis answered that question here: http://forum.mikrotik.com/viewtopic.php?f=7&t=32549

The R52n does not support turbo mode. I don't think that is in the 802.11n draft specifications.
I don't think it was in 802.11a/g either but I seem to remember that atheros added Turbo on their own outside of the standards.

Tom

Does anyone know this for sure.
Just because ROS dont support it, dont mean the chipset dont support it.

Brian.

Houge · Sat Jun 13, 2009 1:14 pm

Does someone have any answers to this, or some data they made on only 1 rx/tx chain?
This was answered before. You need to connect TWO antennas to the N radio card, not just one.
If you only have one antenna on each side, use the R5H in turbo mode instead.

You do not NEED to connect two antennas.
You do need to connect two for max transfer speed, but thats not what im asking about.

Brian.

Houge · Sat Jun 13, 2009 1:17 pm

We would like to share some of our findings using the new R52N cards.

The setup is RB300 and RB600 with ROS4beta3, 1 dishantenna at each point.

For about 2 years now we have used R52H cards, and gotten a good speed ~25mbit TCP on a 4km link. (2.4GHz turbomode)
Last week we got 2x R52N cards, and hoped for a better link and speeds. But its very dissapointing on 2.4GHz B/G/N performance .
On R52H we used turbomode (40mhz), but the new ones don´t seem to be able to work at 40MHz on 1 rx/tx chain.

Why is it not possible to use 40MHz channels on R52N, like you can on R52H?
Its locked at max 65Mbit when using 1 tx/rx chain, and we tryed alot of settings. (local lab test)
The HT extension channel does absolutely nothing when you use 1 rx/tx chain. Is it a bug?

And could someone explain what the difference between 2.4Ghz b/g/n and 2.4Ghz-only-N is, as I see absolutely no difference, it still uses the same B rates when in only-N mode.

Another odd thing is that it seems it only uses long guard interval 800ns, since the rate is at 65Mbit, and not 72.2Mbit.
How do you change it to short 400ns? You can only change from any to long, and that does not change anything.

With the N card in 2.4GHz B/G/N mode we are now only getting MAX 10Mbit on our link, compaired to ~25Mbit in 2.4Ghz turbomode on R52H. At this point, it seems to be a total waste of money.

And 10Mbit is only if you force it to a higher transfer rate. If the data rate is in default, we are getting no more than 5Mbit. It seems like the ROS dont higher the transfer rate, it keeps switching from 1,2,5.5,11mbit. We got about 30dB SNR on the link.

The only positive thing is a 5dB higher signal strenth and a 10-12dB lower noice floor, which makes us wonder even more, why the transfer speed are so bad, compaired to our old R52H in 2,4Ghz turbomode. The total SNR is 5-10dB better then our R52H, but speeds are 15Mbit lower. That dont make any sense.

Does someone have any answers to this, or some data they made on only 1 rx/tx chain?

Regards
Brian.

To clear up some misunderstandings, let me refrase.

We are not asking for Turbo-G mode on the N cards, we ask why the HT extension channel option doesn't work on 1 rx/tx chain, as the standard says it should be possible to use channel bonding (40Mhz width) with 1 Spatial Stream. which should archive a link rate of 135/150Mbit (with 800/400ns guard interval)

https://www.cwnp.com/cwna/wiley_sg_80211n.pdf

Regards
Brian

avantwireless · Mon Jun 15, 2009 5:23 am

Well, so far I haven't seen _any_ results for real world outside connections longer than a mile (1600M) so I'll throw out our first results. Maybe I'll get a license. ( not likely ). This is a test with sr71a on both sides 2x2 mimo w/ 24db 5.8 ghz panels and Rb433 on one side and RB433AH on the AP side. The distance is: 5.55 Miles. What we have seen is a max of 48mb/s udp and 19mb/s tcp one way with btest.exe or using the onboard bandwidth test. We tested with 20mhz channels with and without above or below extension channels. Antenna spacing is 3' on one side and 2' on the other. All AMPDU priorities are checked and we are using the 1st and 3rd chain. Signal levels are in the 55's both ways. So this is no amazing connection. I was hoping to see at least 25mb/s tcp and wishing for 35mb/s. Muchly disappointed but not surprised. Please throw as much idea stuff against the wall and I'm willing to throw it into the setup to try.

martinsdx · Mon Jun 15, 2009 11:44 am

Is it possible to use R52n cards (even without n) on ROS v3.x ?

Mon Jun 15, 2009 11:48 am

Is it possible to use R52n cards (even without n) on ROS v3.x ?

No. They do not appear as interface until you upgrade to 4.x.

martinsdx · Mon Jun 15, 2009 11:52 am

It shows on system/resources/pci as unknown device. Is it possible to get a driver?

RK · Mon Jun 15, 2009 4:48 pm

What we have seen is a max of 48mb/s udp and 19mb/s tcp one way with btest.exe or using the onboard bandwidth test.

We get more than 19 Mbps TCP with just regular 802.11a at 20 MHz, this does not look right.
Maybe try different channels or nstreme?

Mon Jun 15, 2009 5:18 pm

Well, so far I haven't seen _any_ results for real world outside connections longer than a mile (1600M) so I'll throw out our first results. Maybe I'll get a license. ( not likely ). This is a test with sr71a on both sides 2x2 mimo w/ 24db 5.8 ghz panels and Rb433 on one side and RB433AH on the AP side. The distance is: 5.55 Miles. What we have seen is a max of 48mb/s udp and 19mb/s tcp one way with btest.exe or using the onboard bandwidth test. We tested with 20mhz channels with and without above or below extension channels. Antenna spacing is 3' on one side and 2' on the other. All AMPDU priorities are checked and we are using the 1st and 3rd chain. Signal levels are in the 55's both ways. So this is no amazing connection. I was hoping to see at least 25mb/s tcp and wishing for 35mb/s. Muchly disappointed but not surprised. Please throw as much idea stuff against the wall and I'm willing to throw it into the setup to try.

The Antennas are different in polarization?
You've checked with the same setup in lab?
Interference?
Have you tried to reduce tx-power?
I've seen sometimes the extension channel does not work. Reboot helped.

Stefan

infowest · Tue Jun 16, 2009 11:04 pm

All AMPDU priorities are checked and we are using the 1st and 3rd chain.

I think this is your problem.

The chains are 0,1,2,3

Are you using the FIRST and THIRD chain (0 and 2) or 1 and 3? You should probably be using 0 and 2 if you are using the two outside antenna connectors.

avantwireless · Tue Jun 16, 2009 11:08 pm

We have the first and last boxes checked out of three choices...

Same setup in lab ran 100mb/s udp and 55mb/s tcp all tests are done through the radios to pc's on both sides.

netikelis · Wed Jun 17, 2009 1:44 am

Hi All ,
I tested ROS4b3 on RB411AH with SR71-A UBNT cards in only 5G-N mode. All work fine: real file transfer TCP rate ~80Mgps. HT extention channel off. Nstreeme on. Signal on both sides ~ -60dbm. Configuration AP bridge > Station WDS.
Only one problem: randomly "disconected, extensive data loss". Absolutely clear space , no interferences !

log.JPG

omega-00 · Wed Jun 17, 2009 1:49 am

Someones got their own real world link up and running - http://forum.mikrotik.com/viewtopic.php?f=7&t=32523

Wed Jun 17, 2009 8:02 am

Hi All ,
I tested ROS4b3 on RB411AH with SR71-A UBNT cards in only 5G-N mode. All work fine: real file transfer TCP rate ~80Mgps. HT extention channel off. Nstreeme on. Signal on both sides ~ -60dbm. Configuration AP bridge > Station WDS.
Only one problem: randomly "disconected, extensive data loss". Absolutely clear space , no interferences !
log.JPG

Nstreme is not usable with 11N at the moment. Turn it of and your connection is better.

Stefan

NetTraptor · Fri Jun 19, 2009 10:48 am

Ahha gotcha... Periodic recalibration is Broke. Disable it to stop your links disconnecting.
Here in Athens we got our Feeders production line started and already have 2 links in production and 1 in testing (3, 6 and 10+ Km). We are full on for changing most of our Backbone links to 802.11n

First tests show the following
R52n Cards and X86 systems where used. Tests have been performed with various antennas. All numbers are TCP throughput. I know UDP looks nice but...

α> Two Channels (extended channel mode), Two chains, nstreme = Receive Sustain 164Mbps, Send Sustain 165Mbps, Both Sustain 100Mbps. (full duplex Ethernet LOL) [mcs 0,15]
b> Two Channels (extended channel mode), One chain, nstreme = Receive Sustain 100Mbps, Send Sustain 100Mbps, Both Sustain 55Mbps. [mcs 0,15]
c> One Channel Two chains, nstreme = Receive Sustain 90Mbps, Send Sustain 90Mbps, Both Sustain 50Mbps. [mcs 0,8]
d> One Channel, One chain, nstreme = Receive Sustain 48Mbps, Send Sustain 48Mbps, Both Sustain 25Mbps. [mcs 0,8]

No Nstream gives disappointing results. Also noticed only Best fit works OK which is a shame since Exact Size has Best performance and stability in 802.11a

To cut the long story short Here are the OPTIMUM settings for 802.11n. Those might change slightly with time as Nstream support will be added, recalibration should be fixed and other erratic behavior of the driver will be ironed out

/interface wireless security-profiles
set default authentication-types="" eap-methods=passthrough group-ciphers="" \
    group-key-update=5m interim-update=0s management-protection=disabled \
    management-protection-key="" mode=none name=default \
    radius-eap-accounting=no radius-mac-accounting=no \
    radius-mac-authentication=no radius-mac-caching=disabled \
    radius-mac-format=XX:XX:XX:XX:XX:XX radius-mac-mode=as-username \
    static-algo-0=none static-algo-1=none static-algo-2=none static-algo-3=\
    none static-key-0="" static-key-1="" static-key-2="" static-key-3="" \
    static-sta-private-algo=none static-sta-private-key="" \
    static-transmit-key=key-0 supplicant-identity=MikroTik tls-certificate=\
    none tls-mode=no-certificates unicast-ciphers="" wpa-pre-shared-key="" \
    wpa2-pre-shared-key=""
/interface wireless
set 0 ack-timeout=dynamic adaptive-noise-immunity=none allow-sharedkey=no \
    antenna-gain=0 antenna-mode=ant-a area="" arp=enabled band=5ghz-onlyn \
    basic-rates-a/g="" basic-rates-b="" comment="" compression=no country=\
    no_country_set default-ap-tx-limit=0 default-authentication=yes \
    default-client-tx-limit=0 default-forwarding=yes dfs-mode=none \
    disable-running-check=no disabled=no disconnect-timeout=3s \
    frame-lifetime=0 frequency=5180 frequency-mode=manual-txpower hide-ssid=\
    no ht-ampdu-priorities=0,1,2,3,4,5,6,7 ht-amsdu-limit=8192 \
    ht-amsdu-threshold=8192 ht-basic-mcs=mcs-0 ht-extension-channel=disabled \
    ht-guard-interval=any ht-rxchains=0,1 ht-supported-mcs=\
    mcs-0,mcs-7,mcs-8,mcs-14,mcs-15 ht-txchains=0,1 \
    hw-fragmentation-threshold=disabled hw-protection-mode=none \
    hw-protection-threshold=0 hw-retries=15 mac-address=00:0C:42:3A:CF:E9 \
    max-station-count=2 mode=bridge mtu=1500 name=wlan1 on-fail-retry-time=\
    100ms periodic-calibration=disabled periodic-calibration-interval=60 \
    preamble-mode=short proprietary-extensions=post-2.9.25 radio-name=\
    000C423ACFE9 rate-set=configured scan-list=5180,5760,5320,5825 \
    security-profile=default ssid=awmn-2315-3665 station-bridge-clone-mac=\
    00:00:00:00:00:00 supported-rates-a/g="" supported-rates-b="" tx-power=3 \
    tx-power-mode=all-rates-fixed update-stats-interval=disabled \
    wds-cost-range=50-150 wds-default-bridge=none wds-default-cost=100 \
    wds-ignore-ssid=no wds-mode=disabled wmm-support=enabled
/interface wireless manual-tx-power-table
set wlan1 comment="" manual-tx-powers="1Mbps:17,2Mbps:17,5.5Mbps:17,11Mbps:17,\
    6Mbps:17,9Mbps:17,12Mbps:17,18Mbps:17,24Mbps:17,36Mbps:17,48Mbps:17,54Mbps\
    :17,HT20-1:0,HT20-2:0,HT20-3:0,HT20-4:0,HT20-5:0,HT20-6:0,HT20-7:0,HT20-8:\
    0,HT40-1:0,HT40-2:0,HT40-3:0,HT40-4:0,HT40-5:0,HT40-6:0,HT40-7:0,HT40-8:0"
/interface wireless nstreme
set wlan1 comment="" disable-csma=no enable-nstreme=yes enable-polling=yes \
    framer-limit=3998 framer-policy=best-fit
/interface wireless align
set active-mode=yes audio-max=-20 audio-min=-100 audio-monitor=\
    00:00:00:00:00:00 filter-mac=00:00:00:00:00:00 frame-size=300 \
    frames-per-second=25 receive-all=no ssid-all=no
/interface wireless sniffer
set channel-time=200ms file-limit=10 file-name="" memory-limit=10 \
    multiple-channels=no only-headers=no receive-errors=no streaming-enabled=\
    no streaming-max-rate=0 streaming-server=0.0.0.0
/interface wireless snooper
set channel-time=200ms multiple-channels=yes receive-errors=no

P.S. Thanks Ysam for your structured collection of BW Test and comments here

magnavox · Fri Jun 19, 2009 3:24 pm

Two Channels?

NetTraptor · Sat Jun 20, 2009 3:32 pm

2 channels = Extended channel mode. It means just that 2 channels. Instead of 20MHz you use 40MHz. The setting that MT has is above or below ( another extra 20MHz) the Base Frequency you used (which already occupies 20MHz).

I edit it to make it clear.

magnavox · Sat Jun 20, 2009 4:21 pm

2 channels = Extended channel mode. It means just that 2 channels. Instead of 20MHz you use 40MHz. The setting that MT has is above or below ( another extra 20MHz) the Base Frequency you used (which already occupies 20MHz).

I edit it to make it clear.

KimC · Sun Jun 21, 2009 9:02 am

How do you get max throughput on a transparent bridge ?

Hardware: 2 x RB600 with UBNT SR71-A
Test tool: 2 x RB1000 as "RB1000 - RB600 - RB600 - RB1000" - all bridged together in one subnet.
All tests have been made indoor with 2x2 Mimo config. Signal levels adjusted to approx. -50 dB, which gives max airspeed at both interfaces.

The best results I have been able to get is when I configure the radios as plain AP/bridge. The L2-bridge is created as eoip. In this case I get 50+ Mbps TCP fdx (as 50/50, 40/60, 20/80, 0/100 or any other). All 4 units are running at 100 % cpu load in this situation, and are probably the limiting factor.

Any other configuration (with/without nstreme, wds, etc) I have tested, yields substantially less, and are not stable.

avantwireless · Sun Jun 21, 2009 6:37 pm

[quote="NetTraptor"]Ahha gotcha... Periodic recalibration is Broke. Disable it to stop your links disconnecting.
Here in Athens we got our Feeders production line started and already have 2 links in production and 1 in testing (3, 6 and 10+ Km). We are full on for changing most of our Backbone links to 802.11n

First tests show the following
R52n Cards and X86 systems where used. Tests have been performed with various antennas. All numbers are TCP throughput. I know UDP looks nice but...

α> Two Channels (extended channel mode), Two chains, nstreme = Receive Sustain 164Mbps, Send Sustain 165Mbps, Both Sustain 100Mbps. (full duplex Ethernet LOL) [mcs 0,15]
b> Two Channels (extended channel mode), One chain, nstreme = Receive Sustain 100Mbps, Send Sustain 100Mbps, Both Sustain 55Mbps. [mcs 0,15]
c> One Channel Two chains, nstreme = Receive Sustain 90Mbps, Send Sustain 90Mbps, Both Sustain 50Mbps. [mcs 0,8]
d> One Channel, One chain, nstreme = Receive Sustain 48Mbps, Send Sustain 48Mbps, Both Sustain 25Mbps. [mcs 0,8]

No Nstream gives disappointing results. Also noticed only Best fit works OK which is a shame since Exact Size has Best performance and stability in 802.11a

To cut the long story short Here are the OPTIMUM settings for 802.11n. Those might change slightly with time as Nstream support will be added, recalibration should be fixed and other erratic behavior of the driver will be ironed out

/interface wireless security-profiles
set default authentication-types="" eap-methods=passthrough group-ciphers="" \
    group-key-update=5m interim-update=0s management-protection=disabled \
    management-protection-key="" mode=none name=default \
    radius-eap-accounting=no radius-mac-accounting=no \
    radius-mac-authentication=no radius-mac-caching=disabled \
    radius-mac-format=XX:XX:XX:XX:XX:XX radius-mac-mode=as-username \
    static-algo-0=none static-algo-1=none static-algo-2=none static-algo-3=\
    none static-key-0="" static-key-1="" static-key-2="" static-key-3="" \
    static-sta-private-algo=none static-sta-private-key="" \
    static-transmit-key=key-0 supplicant-identity=MikroTik tls-certificate=\
    none tls-mode=no-certificates unicast-ciphers="" wpa-pre-shared-key="" \
    wpa2-pre-shared-key=""
/interface wireless
set 0 ack-timeout=dynamic adaptive-noise-immunity=none allow-sharedkey=no \
    antenna-gain=0 antenna-mode=ant-a area="" arp=enabled band=5ghz-onlyn \
    basic-rates-a/g="" basic-rates-b="" comment="" compression=no country=\
    no_country_set default-ap-tx-limit=0 default-authentication=yes \
    default-client-tx-limit=0 default-forwarding=yes dfs-mode=none \
    disable-running-check=no disabled=no disconnect-timeout=3s \
    frame-lifetime=0 frequency=5180 frequency-mode=manual-txpower hide-ssid=\
    no ht-ampdu-priorities=0,1,2,3,4,5,6,7 ht-amsdu-limit=8192 \
    ht-amsdu-threshold=8192 ht-basic-mcs=mcs-0 ht-extension-channel=disabled \
    ht-guard-interval=any ht-rxchains=0,1 ht-supported-mcs=\
    mcs-0,mcs-7,mcs-8,mcs-14,mcs-15 ht-txchains=0,1 \
    hw-fragmentation-threshold=disabled hw-protection-mode=none \
    hw-protection-threshold=0 hw-retries=15 mac-address=00:0C:42:3A:CF:E9 \
    max-station-count=2 mode=bridge mtu=1500 name=wlan1 on-fail-retry-time=\
    100ms periodic-calibration=disabled periodic-calibration-interval=60 \
    preamble-mode=short proprietary-extensions=post-2.9.25 radio-name=\
    000C423ACFE9 rate-set=configured scan-list=5180,5760,5320,5825 \
    security-profile=default ssid=awmn-2315-3665 station-bridge-clone-mac=\
    00:00:00:00:00:00 supported-rates-a/g="" supported-rates-b="" tx-power=3 \
    tx-power-mode=all-rates-fixed update-stats-interval=disabled \
    wds-cost-range=50-150 wds-default-bridge=none wds-default-cost=100 \
    wds-ignore-ssid=no wds-mode=disabled wmm-support=enabled
/interface wireless manual-tx-power-table
set wlan1 comment="" manual-tx-powers="1Mbps:17,2Mbps:17,5.5Mbps:17,11Mbps:17,\
    6Mbps:17,9Mbps:17,12Mbps:17,18Mbps:17,24Mbps:17,36Mbps:17,48Mbps:17,54Mbps\
    :17,HT20-1:0,HT20-2:0,HT20-3:0,HT20-4:0,HT20-5:0,HT20-6:0,HT20-7:0,HT20-8:\
    0,HT40-1:0,HT40-2:0,HT40-3:0,HT40-4:0,HT40-5:0,HT40-6:0,HT40-7:0,HT40-8:0"
/interface wireless nstreme
set wlan1 comment="" disable-csma=no enable-nstreme=yes enable-polling=yes \
    framer-limit=3998 framer-policy=best-fit
/interface wireless align
set active-mode=yes audio-max=-20 audio-min=-100 audio-monitor=\
    00:00:00:00:00:00 filter-mac=00:00:00:00:00:00 frame-size=300 \
    frames-per-second=25 receive-all=no ssid-all=no
/interface wireless sniffer
set channel-time=200ms file-limit=10 file-name="" memory-limit=10 \
    multiple-channels=no only-headers=no receive-errors=no streaming-enabled=\
    no streaming-max-rate=0 streaming-server=0.0.0.0
/interface wireless snooper
set channel-time=200ms multiple-channels=yes receive-errors=no

May I ask your antenna configuration? Spacings and such

martinsdx · Mon Jun 22, 2009 5:48 pm

Never seen anything better than this

(Tested on RB133c3 with R52n and RB411AH with R52n). Bandwith test shows up to 25 - 30 Mbps, not more ...

dingsingo · Mon Jun 22, 2009 6:13 pm

Never seen anything better than this (Tested on RB133c3 with R52n and RB411AH with R52n). Bandwith test shows up to 25 - 30 Mbps, not more ...

Hi,

sorry SNR is 7 , not more ???

Thx

dingsingo

rebel2234 · Mon Jun 22, 2009 8:30 pm

I would say with the 00:15:6D mac prefix on the card it is indeed a UBNT card.

Nstreme enabled!!!
Big!

Are you running a UBNT card... SR71-A?
Link distance?
Number of antennas: x2 or x3?

jone8880 · Tue Jun 23, 2009 1:03 am

i was make a big mistake.... after succesfull test "on the table"( 200 Mbs wow!)) today i bring it to working link for replace nstream dual.(50+ 50 mbps)..distance 3 km.. 5G....2 antennas with differente polarisation... no noise in air....

link working perfect 2 minutes..( 270 HT/270 HT)// 95-100% quality. 90+90 Mbps...but after...quality reduce to 10-15%....only 20 Mbs in two way..
possible to fix it by disable- enable wireless interface and get full working link again for two minutes.

ps. 20 mbs only in nstream mode... without nstream - 7-8 Mbs

NetTraptor · Tue Jun 23, 2009 3:31 am

May I ask your antenna configuration? Spacings and such

We got a guy that does our dual polarity feeders. Still testing configs but we are satisfied with a certain design that we are going to run a few batches.
We tested those with 80 and 100cm sat dishes. I guess a dual polarity Pac can do the job as well.

tgrand · Tue Jun 23, 2009 5:17 am

Ok.

In all the ranter somebodies request when unheard.
regarding the R5(2)N card not being recognized by the Router OS V4beta3.

I also have encountered this.
I have yet to see a real R52N card work.

It it beyond comprehension that some cards are recognized and some are not.

Is it a batch problem?
If so, can I get mine exchanged?

omega-00 · Tue Jun 23, 2009 9:37 am

Have you actually upgraded to 4.03beta and THEN upgraded your licence key as well?

I've detailed how to go about it here: http://wiki.mikrotik.com/wiki/802.11n_S ... ng_802.11n

priit · Tue Jun 23, 2009 10:29 am

this should not be true, all Level4 licenses allow upgrade to v4.xx if they were purchased when v3 was already released.

if not, the upgraded key will fix it (when you upgrade to v4beta, it makes a new key, and makes sure that your upgrade times are correct).

Well, this is possible only when using v3.25
Most of the early RB600 we have here are Level 4 and showing upgradability only to 3.x
As RB600 runs only v3 and up, I can conclude that it was purchased when v3 was released

Starting at v3.25 there appears the "Update License Key" tab and after the procedure the v4 will be shown as possible.

rgds,
Priit

phendry · Tue Jun 23, 2009 12:56 pm

I have yet to see a real R52N card work.

Received our R52N's just before the weekend. Upgraded a couple of RB411AH's to V4 and upgraded the license as per the instructions and all showed up fine. Couple of quick tests in the lab showed all working as per design

tgrand · Tue Jun 23, 2009 3:31 pm

Thanks Omega.

Bizarre that one would have to upgrade the license even though it showed Level 4 as the license.
The only noticable change was that the R52N is now being recognized.

Thanks again.

avantwireless · Tue Jun 23, 2009 4:33 pm

Thanks Omega.

Bizarre that one would have to upgrade the license even though it showed Level 4 as the license.
The only noticable change was that the R52N is now being recognized.

Thanks again.

Ummmm maybe you are confusing Level 4 with Version 4? Levels are the set of features that are available in a Version. Version 4 is required to run any N products. Version 4 is the Beta that we are all testing. Level 4 is the Wireless AP license that MT has made the required License level to get an automatic pass for upgrading to the Beta Version.

tgrand · Tue Jun 23, 2009 5:56 pm

The version was at V4beta3
The license level was at Level 4.

Still applying the upgrade to the license, did not change the above information, yet the R52H was then discovered.
No Confusion, just strangeness.

infowest · Tue Jun 23, 2009 8:35 pm

Thanks for all the tips on getting these working better. We've followed many of them, here are the ones that seem to have helped us so far:

- No WDS
- Short preamble
- No Periodic calibration
- Nstreme on, but CSMA not disabled
- Nstreme best fit
- Play with power levels until we get best results. Seems pretty picky on this.
- Be sure the routerboard firmware (not just RouterOS) is the latest revision ( /system routerboard print and /system routerboard upgrade)

Does anyone know if adaptive noise immunity does anything on these cards (R52N in 5GHZ-N mode)?

Any other tips?

avantwireless · Wed Jun 24, 2009 7:02 am

Thanks for all the tips on getting these working better. We've followed many of them, here are the ones that seem to have helped us so far:

- No WDS
- Short preamble
- No Periodic calibration
- Nstreme on, but CSMA not disabled
- Nstreme best fit
- Play with power levels until we get best results. Seems pretty picky on this.
- Be sure the routerboard firmware (not just RouterOS) is the latest revision ( /system routerboard print and /system routerboard upgrade)

Does anyone know if adaptive noise immunity does anything on these cards (R52N in 5GHZ-N mode)?

Any other tips?

Thank you, this was a huge help. With this we got 32mb/s tcp in a very noisy urban area on a 5.5 mile link with sr71a and two 24db panels at either end. This is a 2x improvement over what we were seeing and finally fast enough to justify using it. One end was a 433ah and the other just a 433. The 433 was maxed out at 100% with 32mb/s flowing through it so we should see a nice improvement when we upgrade that to a 433ah. As always the view is a lot better when you can get on the shoulders of others. Thank you and all the others that contributed..!

dingsingo · Wed Jun 24, 2009 2:24 pm

Hi,

my Testsystem over 8km will make not more the 54Mbit brutto. Why the System dosn't go to HT-Mode ?
RB433 V4.beta3 with ~30dB Signal

The Config form AP ist:

/interface wireless
set 0 ack-timeout=dynamic adaptive-noise-immunity=none allow-sharedkey=no \
antenna-gain=0 antenna-mode=ant-a area="" arp=enabled band=5ghz-onlyn \
basic-rates-a/g=6Mbps basic-rates-b=1Mbps comment="" compression=no \
country=germany default-ap-tx-limit=0 default-authentication=yes \
default-client-tx-limit=0 default-forwarding=yes dfs-mode=no-radar-detect \
disable-running-check=no disabled=no disconnect-timeout=3s \
frame-lifetime=0 frequency=5180 frequency-mode=regulatory-domain \
hide-ssid=no ht-ampdu-priorities=0,1,2,3,4,5,6,7 ht-amsdu-limit=8192 \
ht-amsdu-threshold=8192 ht-basic-mcs=\
mcs-0,mcs-1,mcs-2,mcs-3,mcs-4,mcs-5,mcs-6,mcs-7 ht-extension-channel=\
disabled ht-guard-interval=any ht-rxchains=0,1 ht-supported-mcs="mcs-0,mcs\
-1,mcs-2,mcs-3,mcs-4,mcs-5,mcs-6,mcs-7,mcs-8,mcs-9,mcs-10,mcs-11,mcs-12,mc\
s-13,mcs-14,mcs-15" ht-txchains=0,1 hw-fragmentation-threshold=disabled \
hw-protection-mode=none hw-protection-threshold=0 hw-retries=4 \
mac-address=00:0C:42:3A:C7:30 max-station-count=2007 mode=ap-bridge mtu=\
1500 name=wlan1 on-fail-retry-time=100ms periodic-calibration=default \
periodic-calibration-interval=60 preamble-mode=both \
proprietary-extensions=post-2.9.25 radio-name="test olf" rate-set=default \
scan-list=5540,5560 security-profile=WPA2 ssid=89 \
station-bridge-clone-mac=00:00:00:00:00:00 supported-rates-a/g=\
6Mbps,9Mbps,12Mbps,18Mbps,24Mbps,36Mbps,48Mbps,54Mbps supported-rates-b=\
1Mbps,2Mbps,5.5Mbps,11Mbps tx-power-mode=default update-stats-interval=\
disabled wds-cost-range=50-150 wds-default-bridge=none wds-default-cost=\
100 wds-ignore-ssid=no wds-mode=disabled wmm-support=enabled
/interface wireless manual-tx-power-table
set wlan1 comment="" manual-tx-powers="1Mbps:17,2Mbps:17,5.5Mbps:17,11Mbps:17,\
6Mbps:17,9Mbps:17,12Mbps:17,18Mbps:17,24Mbps:17,36Mbps:17,48Mbps:17,54Mbps\
:17,HT20-1:0,HT20-2:0,HT20-3:0,HT20-4:0,HT20-5:0,HT20-6:0,HT20-7:0,HT20-8:\
0,HT40-1:0,HT40-2:0,HT40-3:0,HT40-4:0,HT40-5:0,HT40-6:0,HT40-7:0,HT40-8:0"
/interface wireless nstreme
set wlan1 comment="" disable-csma=no enable-nstreme=no enable-polling=yes \
framer-limit=3200 framer-policy=best-fit
/interface wireless align
set active-mode=yes audio-max=-20 audio-min=-100 audio-monitor=\
00:00:00:00:00:00 filter-mac=00:00:00:00:00:00 frame-size=300 \
frames-per-second=25 receive-all=no ssid-all=no

Client :

/interface wireless
set 0 ack-timeout=dynamic adaptive-noise-immunity=none allow-sharedkey=no antenna-gain=0 antenna-mode=ant-a area="" arp=enabled \
band=5ghz-onlyn basic-rates-a/g=6Mbps basic-rates-b=1Mbps comment="" compression=no country=germany default-ap-tx-limit=0 \
default-authentication=yes default-client-tx-limit=0 default-forwarding=yes dfs-mode=no-radar-detect disable-running-check=no \
disabled=no disconnect-timeout=3s frame-lifetime=0 frequency=5180 frequency-mode=regulatory-domain hide-ssid=no \
ht-ampdu-priorities=0,1,2,3,4,5,6,7 ht-amsdu-limit=8192 ht-amsdu-threshold=8192 ht-basic-mcs=\
mcs-0,mcs-1,mcs-2,mcs-3,mcs-4,mcs-5,mcs-6,mcs-7 ht-extension-channel=disabled ht-guard-interval=any ht-rxchains=0,1 \
ht-supported-mcs=mcs-0,mcs-1,mcs-2,mcs-3,mcs-4,mcs-5,mcs-6,mcs-7,mcs-8,mcs-9,mcs-10,mcs-11,mcs-12,mcs-13,mcs-14,mcs-15 \
ht-txchains=0,1 hw-fragmentation-threshold=disabled hw-protection-mode=none hw-protection-threshold=0 hw-retries=4 mac-address=\
00:0C:42:3A:C7:7F max-station-count=2007 mode=station mtu=1600 name=wlan1 on-fail-retry-time=100ms periodic-calibration=default \
periodic-calibration-interval=60 preamble-mode=both proprietary-extensions=post-2.9.25 radio-name=000C423AC77F rate-set=default \
scan-list=default security-profile=WPA2 ssid=89 station-bridge-clone-mac=00:00:00:00:00:00 supported-rates-a/g=\
6Mbps,9Mbps,12Mbps,18Mbps,24Mbps,36Mbps,48Mbps,54Mbps supported-rates-b=1Mbps,2Mbps,5.5Mbps,11Mbps tx-power-mode=default \
update-stats-interval=disabled wds-cost-range=50-150 wds-default-bridge=none wds-default-cost=100 wds-ignore-ssid=no wds-mode=\
disabled wmm-support=disabled
/interface wireless manual-tx-power-table
set wlan1 comment="" manual-tx-powers="1Mbps:17,2Mbps:17,5.5Mbps:17,11Mbps:17,6Mbps:17,9Mbps:17,12Mbps:17,18Mbps:17,24Mbps:17,36Mbps:\
17,48Mbps:17,54Mbps:17,HT20-1:0,HT20-2:0,HT20-3:0,HT20-4:0,HT20-5:0,HT20-6:0,HT20-7:0,HT20-8:0,HT40-1:0,HT40-2:0,HT40-3:0,HT40-4:\
0,HT40-5:0,HT40-6:0,HT40-7:0,HT40-8:0"
/interface wireless nstreme
set wlan1 comment="" disable-csma=no enable-nstreme=no enable-polling=yes framer-limit=3200 framer-policy=none
/interface wireless align
set active-mode=yes audio-max=-20 audio-min=-100 audio-monitor=00:00:00:00:00:00 filter-mac=00:00:00:00:00:00 frame-size=300 \
frames-per-second=25 receive-all=no ssid-all=no
/interface wireless sniffer
set channel-time=200ms file-limit=10 file-name="" memory-limit=10 multiple-channels=no only-headers=no receive-errors=no \
streaming-enabled=no streaming-max-rate=0 streaming-server=0.0.0.0
/interface wireless snooper
set channel-time=200ms multiple-channels=yes receive-errors=no

Thy

dingsingo

fly · Wed Jun 24, 2009 3:04 pm

Hi,

my Testsystem over 8km will make not more the 54Mbit brutto. Why the System dosn't go to HT-Mode ?
RB433 V4.beta3 with ~30dB Signal

The Config form AP ist:

/interface wireless
set 0 ack-timeout=dynamic adaptive-noise-immunity=none allow-sharedkey=no \
antenna-gain=0 antenna-mode=ant-a area="" arp=enabled band=5ghz-onlyn \
basic-rates-a/g=6Mbps basic-rates-b=1Mbps comment="" compression=no \
country=germany default-ap-tx-limit=0 default-authentication=yes \
default-client-tx-limit=0 default-forwarding=yes dfs-mode=no-radar-detect \
disable-running-check=no disabled=no disconnect-timeout=3s \
frame-lifetime=0 frequency=5180 frequency-mode=regulatory-domain \
hide-ssid=no ht-ampdu-priorities=0,1,2,3,4,5,6,7 ht-amsdu-limit=8192 \
ht-amsdu-threshold=8192 ht-basic-mcs=\
mcs-0,mcs-1,mcs-2,mcs-3,mcs-4,mcs-5,mcs-6,mcs-7 ht-extension-channel=\
disabled ht-guard-interval=any ht-rxchains=0,1 ht-supported-mcs="mcs-0,mcs\
-1,mcs-2,mcs-3,mcs-4,mcs-5,mcs-6,mcs-7,mcs-8,mcs-9,mcs-10,mcs-11,mcs-12,mc\
s-13,mcs-14,mcs-15" ht-txchains=0,1 hw-fragmentation-threshold=disabled \
hw-protection-mode=none hw-protection-threshold=0 hw-retries=4 \
mac-address=00:0C:42:3A:C7:30 max-station-count=2007 mode=ap-bridge mtu=\
1500 name=wlan1 on-fail-retry-time=100ms periodic-calibration=default \
periodic-calibration-interval=60 preamble-mode=both \
proprietary-extensions=post-2.9.25 radio-name="test olf" rate-set=default \
scan-list=5540,5560 security-profile=WPA2 ssid=89 \
station-bridge-clone-mac=00:00:00:00:00:00 supported-rates-a/g=\
6Mbps,9Mbps,12Mbps,18Mbps,24Mbps,36Mbps,48Mbps,54Mbps supported-rates-b=\
1Mbps,2Mbps,5.5Mbps,11Mbps tx-power-mode=default update-stats-interval=\
disabled wds-cost-range=50-150 wds-default-bridge=none wds-default-cost=\
100 wds-ignore-ssid=no wds-mode=disabled wmm-support=enabled
/interface wireless manual-tx-power-table
set wlan1 comment="" manual-tx-powers="1Mbps:17,2Mbps:17,5.5Mbps:17,11Mbps:17,\
6Mbps:17,9Mbps:17,12Mbps:17,18Mbps:17,24Mbps:17,36Mbps:17,48Mbps:17,54Mbps\
:17,HT20-1:0,HT20-2:0,HT20-3:0,HT20-4:0,HT20-5:0,HT20-6:0,HT20-7:0,HT20-8:\
0,HT40-1:0,HT40-2:0,HT40-3:0,HT40-4:0,HT40-5:0,HT40-6:0,HT40-7:0,HT40-8:0"
/interface wireless nstreme
set wlan1 comment="" disable-csma=no enable-nstreme=no enable-polling=yes \
framer-limit=3200 framer-policy=best-fit
/interface wireless align
set active-mode=yes audio-max=-20 audio-min=-100 audio-monitor=\
00:00:00:00:00:00 filter-mac=00:00:00:00:00:00 frame-size=300 \
frames-per-second=25 receive-all=no ssid-all=no

Client :

/interface wireless
set 0 ack-timeout=dynamic adaptive-noise-immunity=none allow-sharedkey=no antenna-gain=0 antenna-mode=ant-a area="" arp=enabled \
band=5ghz-onlyn basic-rates-a/g=6Mbps basic-rates-b=1Mbps comment="" compression=no country=germany default-ap-tx-limit=0 \
default-authentication=yes default-client-tx-limit=0 default-forwarding=yes dfs-mode=no-radar-detect disable-running-check=no \
disabled=no disconnect-timeout=3s frame-lifetime=0 frequency=5180 frequency-mode=regulatory-domain hide-ssid=no \
ht-ampdu-priorities=0,1,2,3,4,5,6,7 ht-amsdu-limit=8192 ht-amsdu-threshold=8192 ht-basic-mcs=\
mcs-0,mcs-1,mcs-2,mcs-3,mcs-4,mcs-5,mcs-6,mcs-7 ht-extension-channel=disabled ht-guard-interval=any ht-rxchains=0,1 \
ht-supported-mcs=mcs-0,mcs-1,mcs-2,mcs-3,mcs-4,mcs-5,mcs-6,mcs-7,mcs-8,mcs-9,mcs-10,mcs-11,mcs-12,mcs-13,mcs-14,mcs-15 \
ht-txchains=0,1 hw-fragmentation-threshold=disabled hw-protection-mode=none hw-protection-threshold=0 hw-retries=4 mac-address=\
00:0C:42:3A:C7:7F max-station-count=2007 mode=station mtu=1600 name=wlan1 on-fail-retry-time=100ms periodic-calibration=default \
periodic-calibration-interval=60 preamble-mode=both proprietary-extensions=post-2.9.25 radio-name=000C423AC77F rate-set=default \
scan-list=default security-profile=WPA2 ssid=89 station-bridge-clone-mac=00:00:00:00:00:00 supported-rates-a/g=\
6Mbps,9Mbps,12Mbps,18Mbps,24Mbps,36Mbps,48Mbps,54Mbps supported-rates-b=1Mbps,2Mbps,5.5Mbps,11Mbps tx-power-mode=default \
update-stats-interval=disabled wds-cost-range=50-150 wds-default-bridge=none wds-default-cost=100 wds-ignore-ssid=no wds-mode=\
disabled wmm-support=disabled
/interface wireless manual-tx-power-table
set wlan1 comment="" manual-tx-powers="1Mbps:17,2Mbps:17,5.5Mbps:17,11Mbps:17,6Mbps:17,9Mbps:17,12Mbps:17,18Mbps:17,24Mbps:17,36Mbps:\
17,48Mbps:17,54Mbps:17,HT20-1:0,HT20-2:0,HT20-3:0,HT20-4:0,HT20-5:0,HT20-6:0,HT20-7:0,HT20-8:0,HT40-1:0,HT40-2:0,HT40-3:0,HT40-4:\
0,HT40-5:0,HT40-6:0,HT40-7:0,HT40-8:0"
/interface wireless nstreme
set wlan1 comment="" disable-csma=no enable-nstreme=no enable-polling=yes framer-limit=3200 framer-policy=none
/interface wireless align
set active-mode=yes audio-max=-20 audio-min=-100 audio-monitor=00:00:00:00:00:00 filter-mac=00:00:00:00:00:00 frame-size=300 \
frames-per-second=25 receive-all=no ssid-all=no
/interface wireless sniffer
set channel-time=200ms file-limit=10 file-name="" memory-limit=10 multiple-channels=no only-headers=no receive-errors=no \
streaming-enabled=no streaming-max-rate=0 streaming-server=0.0.0.0
/interface wireless snooper
set channel-time=200ms multiple-channels=yes receive-errors=no

Thy

dingsingo

I had the same issue, I solved removing all supported data-rates a/g on both cards.

fly · Wed Jun 24, 2009 3:17 pm

Someones got their own real world link up and running - http://forum.mikrotik.com/viewtopic.php?f=7&t=32523

9Km, Nstreme enabled, 30Mbps Full Duplex TCP, 14d uptime.

InoX · Wed Jun 24, 2009 4:23 pm

I have 30mbps full duplex with regular 40Mhz channel.
Correction...35mbps not 30.

dingsingo · Wed Jun 24, 2009 5:15 pm

Hi,

ok it works. But 20Mhz and ~30dB Signal. Max. 35Mbit in one way by H20-8. Two-Way ~2Mbit max. on 8km an other 802.11a in the neighborhood CCQ ~60-80%.

dingsingo

merlinthemagic · Wed Jun 24, 2009 11:36 pm

Hey All,

We setup a 6 mile link between 2 high rise buildings using a single RB52n on each side with a 32dBi dual pol antenna.

The Signal strength is -63 when locked at 130Mbits-HT (don't know the mW output of the RB52 at that rate).

The results are mixed, but one major thing stands out, if the radio is allowed to select its own transmit rate the Megabit throughput is in the single digits. If we force the radio to use "MCS 15" we get the result in the pictures below. The tests are done from RB1000's.

Ap config (mirrored on station):

0 R name="wlan1" mtu=1500 mac-address=00:0C:42:3A:DA:AD arp=enabled disable-running-check=no interface-type=Atheros 11N radio-name="000C423ADAAD" mode=ap-bridge ssid="MikroTik1" area=""
frequency-mode=manual-txpower country=no_country_set antenna-gain=0 frequency=5280 band=5ghz-onlyn scan-list=default rate-set=configured supported-rates-b=1Mbps,2Mbps,5.5Mbps,11Mbps
supported-rates-a/g="" basic-rates-b=1Mbps basic-rates-a/g="" max-station-count=2007 ack-timeout=dynamic tx-power-mode=default periodic-calibration=default periodic-calibration-interval=60
dfs-mode=none antenna-mode=ant-a wds-mode=disabled wds-default-bridge=none wds-default-cost=100 wds-cost-range=50-150 wds-ignore-ssid=no update-stats-interval=disabled default-authentication=yes
default-forwarding=yes default-ap-tx-limit=0 default-client-tx-limit=0 proprietary-extensions=post-2.9.25 wmm-support=disabled hide-ssid=no security-profile=default disconnect-timeout=3s
on-fail-retry-time=100ms preamble-mode=both compression=no allow-sharedkey=no station-bridge-clone-mac=00:00:00:00:00:00 ht-ampdu-priorities=0,1,2,3,4,5,6,7 ht-guard-interval=any
ht-extension-channel=disabled ht-supported-mcs=mcs-0,mcs-1,mcs-2,mcs-3,mcs-4,mcs-5,mcs-6,mcs-7,mcs-8,mcs-9,mcs-10,mcs-11,mcs-12,mcs-13,mcs-14,mcs-15 ht-basic-mcs=mcs-15 ht-txchains=0,1
ht-rxchains=0,1 ht-amsdu-limit=8192 ht-amsdu-threshold=8192 hw-retries=4 frame-lifetime=0 adaptive-noise-immunity=none hw-fragmentation-threshold=disabled hw-protection-mode=none
hw-protection-threshold=0

ianek · Thu Jun 25, 2009 9:05 am

Hi,
I made some test - 6km link, 2x RB433 with dual polarization antennas and R52n.
Result for me is , that the rate selection for HT mode is broken, rate is shifting from 6/6 in a mode to 58,5/58,5-HT mode , when locked i can get some improvement, removing a/g rates help a bit, but it is not "final" solution i think.

Jan

dingsingo · Thu Jun 25, 2009 11:39 am

Hi,

my Testlink broke by more than 10Mbit throuput.

The Client say in Log: "lost connection, not polled for too long"

dingsingo

Thu Jun 25, 2009 1:00 pm

i was make a big mistake.... after succesfull test "on the table"( 200 Mbs wow!)) today i bring it to working link for replace nstream dual.(50+ 50 mbps)..distance 3 km.. 5G....2 antennas with differente polarisation... no noise in air....

link working perfect 2 minutes..( 270 HT/270 HT)// 95-100% quality. 90+90 Mbps...but after...quality reduce to 10-15%....only 20 Mbs in two way..
possible to fix it by disable- enable wireless interface and get full working link again for two minutes.

ps. 20 mbs only in nstream mode... without nstream - 7-8 Mbs

Try disabling periodical-calibration and test again.

Thu Jun 25, 2009 1:13 pm

Hi,

my Testlink broke by more than 10Mbit throuput.

The Client say in Log: "lost connection, not polled for too long"

dingsingo

try disabling the periodical calibration and disable nstreme.

dingsingo · Thu Jun 25, 2009 2:00 pm

Hi,

my Testlink broke by more than 10Mbit throuput.

The Client say in Log: "lost connection, not polled for too long"

dingsingo
try disabling the periodical calibration and disable nstreme.

Hi,
I've disable the parameter. But not realy better. I thing, the Problem is, the we have other 5Ghz Link on both Side. So the Band 2 & 3 not realy free. But a normale 5Ghz 20 Mhz Link on my 8km works fine, with ~20Mbit netto. Or 4-7 Mbit in both Side.

dingsingo

fly · Thu Jun 25, 2009 2:41 pm

I have 30mbps full duplex with regular 40Mhz channel.
Correction...35mbps not 30.

It's 37Mbps full duplex, for precision... and I'm using 20Mhz.

dingsingo · Thu Jun 25, 2009 3:10 pm

I have 30mbps full duplex with regular 40Mhz channel.
Correction...35mbps not 30.
It's 37Mbps full duplex, for precision... and I'm using 20Mhz.

Hi, have you other 5Ghz Link in the neighborhood ?

dingsingo

fly · Thu Jun 25, 2009 4:03 pm

I have 30mbps full duplex with regular 40Mhz channel.
Correction...35mbps not 30.
It's 37Mbps full duplex, for precision... and I'm using 20Mhz.

Hi, have you other 5Ghz Link in the neighborhood ?

dingsingo

I installed it on a tower really full of 5GHz links.
Imagine that while scanning I need to use the scrollbar in Winbox!
I use a dual-pol integrated antenna, and signal strenght is really good (-57).
Are you using rf cables or just pigtails as I do?

dingsingo · Thu Jun 25, 2009 5:36 pm

Hi @fly

we use on both Side Dual-Ponyting 20dB with 20cm Pigtails. SNR ~40dB.

dingsingo

mp3turbo · Thu Jun 25, 2009 5:47 pm

uldis, am I right when I say what we know for sure:

- periodic calibration is broken
- there is something wrong with automatic rate selection
- nstreme is not adapted yet to 11n (I'm not saying broken, it just has to be done different way then until now)

is this correct? I'm eagerly waiting for beta4...

jone8880 · Thu Jun 25, 2009 7:53 pm

i was make a big mistake.... after succesfull test "on the table"( 200 Mbs wow!)) today i bring it to working link for replace nstream dual.(50+ 50 mbps)..distance 3 km.. 5G....2 antennas with differente polarisation... no noise in air....

link working perfect 2 minutes..( 270 HT/270 HT)// 95-100% quality. 90+90 Mbps...but after...quality reduce to 10-15%....only 20 Mbs in two way..
possible to fix it by disable- enable wireless interface and get full working link again for two minutes.

ps. 20 mbs only in nstream mode... without nstream - 7-8 Mbs
Try disabling periodical-calibration and test again.

calibration was disabled.. with calibration + nstream i've got disconnection every 20-30 sec

ianek · Thu Jun 25, 2009 10:44 pm

With nstreme enabled i got disconnects almost all the time under load ... periodic calibration off, csma not disabled, framing best fit, tried a lot of combinations, the nstreme with N mode is not stable

Jan

ghmorris · Fri Jun 26, 2009 12:41 am

With nstreme enabled i got disconnects almost all the time under load ... periodic calibration off, csma not disabled, framing best fit, tried a lot of combinations, the nstreme with N mode is not stable
Jan

Nsteme is fine for us.

Try using two separate antennas with some vertical separation instead of dual polarity.

We have 3 R52N links in production, longest is 25kms. Completely solid at 34Mbits TCP with Nsteme on and periodic calibration off using a 20MHz channel. I can't find enough clean spectrum to use 40MHz.

We have had lousy results with dual polarity antennas and Atheros cards historically, that is the first place I would look for problems.

George

edit for clarity

jone8880 · Fri Jun 26, 2009 1:51 am

i've got good stability with manual data rate setting (disabling all exept msc 0,7,8,14,15, it possible make via Winbox if you check Data rate- configured)) + nstream===50+50 mb tcp
but for 2 chains +extension channel is not best... i think because no good Tx/rx CCQ - 80-90 %.

ianek · Fri Jun 26, 2009 10:07 am

With nstreme enabled i got disconnects almost all the time under load ... periodic calibration off, csma not disabled, framing best fit, tried a lot of combinations, the nstreme with N mode is not stable
Jan
Nsteme is fine for us.

Try using two separate antennas with some vertical separation instead of dual polarity.

We have 3 R52N links in production, longest is 25kms. Completely solid at 34Mbits TCP with Nsteme on and periodic calibration off using a 20MHz channel. I can't find enough clean spectrum to use 40MHz.

We have had lousy results with dual polarity antennas and Atheros cards historically, that is the first place I would look for problems.

George

edit for clarity

Well,
thanks for reply, but logically - when i switch to pure A with nstreme, then i get aroung 30Mbits HD with nstreme , when the same link in same conditions is set to N mode then it is disconnecting all the time, without nstreme it cant pass more than 5Mbits - using only 1 chain, 2 chains is behaving almost the same, so the dual pole antenna setup is not the problem.
I will wait until Mikrotik will fix rate selection algoritm and nstreme in N mode for more results.

Jan

stefan@on4sh.be · Fri Jun 26, 2009 1:36 pm

Today we did some speedtests on ubnt's RB71-A mimo cards using a 2*2 configuration.
Connector 1 we connected to a vertical polarised 24 dbi panel antenna.
Connector 3 we connected to a horizontal polarised 24 dbi panel antenna.

The antenna's are mounted on the same pole above each other.
Distance between 2 sites is about 4.5 Km / 2.8 miles both sites are +60m above ground level.
Site1 is running on RB600, site2 is running RB433ah.

TX power is about 10 dbm -4db coaxloss and +24 dbi antenna = 30 dbm, SNR = 40 db.

Speeds are +150 mbit udp transfer.
Later testing real tcp with 2 pc's.

More info and pictures on this link, some more test will follow for R52n cards and also 3*3 outdoor setup.
http://www.wirelessantwerpen.be/demo-n/

Regards,
Stefan

Wirelessantwerpen
on4sh

mp3turbo · Fri Jun 26, 2009 1:47 pm

connector3? Aha, UBNT card... I was wondering because Mikrotik r52n has two connectors only.

jone8880 · Fri Jun 26, 2009 11:14 pm

i was testing noname(AR5418) miniPCIe wireless card with 3 connectors via miniPCI adapter.. but used only 2 connectors and have got result like r52n.

spectrasurf · Sat Jun 27, 2009 6:12 am

Let's say your AP is a RB600+SR71 card, with 3 sector antennas covering the same area (i.e. all 90-degree sectors, all pointing the same direction, but mounted on the tower side-by-side). Two vertical, one horizontal.

If you have regular WiFi laptop clients roaming around, using 802.11G (NOT N!) cards, would you be able to theoretically push more data to them overall?

i.e. if with 802.11g AP and 1 antenna you could get a maximum of 10Mbps to all your clients, could you get a maximum of 30Mbps with 3 antennas?

niksoon · Sat Jun 27, 2009 6:19 pm

New test from RTS-Comuters in Bulgaria 802.11n tested on RB433AH

spectrasurf · Sun Jun 28, 2009 5:34 pm

Can you post your config?

We have similar setup, almost exact same signal, and only get about 25Mbps max.

mp3turbo · Sun Jun 28, 2009 6:31 pm

throughput is not just matter of signal ; throughput mainly depends on interference in the air and several much more important factors than strength of signal alone.

= his config will not help you at all.

spectrasurf · Sun Jun 28, 2009 10:11 pm

On one hand you say the that throughput mainly depends on much more than the signal strength; then in the next breath you say his config won't help at all.
Uhh... ya think? If he has things like MCS settings, ANI settings, TX power settings, or about 10 other settings anything other than default it certainly WILL make a difference.

Your post was a waste of my time, and bandwidth.

As I originally asked, if you could share the config you used to get that throughput, it'd be appreciated.

tgrand · Mon Jun 29, 2009 6:48 am

I see a few postings with ping times; Are these ping times just across the wireless 'N' link?
If so 60-69 ms is terrible.
Keeping latencies down is important also.

What is the deal with the high ping times?

greenieofdubbo · Mon Jun 29, 2009 7:08 am

What is the deal with the high ping times?

My guess is because the links are flooded with speed tests at the time.

masked · Tue Jun 30, 2009 1:58 am

My Mikrotik "n" cards are arriving on Monday so I have not actually tested this yet. But I believe you will not see any improvement over 802.11a with this link. You need at least two antennas at each end for mimo (n).

Tom
MIMO exists to lower effective EVM and other distortions. Technically MIMO is (now, at least in the newer drafts) part of the pre-standard but its not actually required for an 11n link (at least at HT20 rates). It sounds like (BUT I HAVE NO PERSONAL EXPERIENCE YET) it requires 2 antennas to run HT40 rates, but that's a limitation of Atheros's chipset, not the protocol itself.

I'll be ordering some 11n cards from you in the next few days. I want to test the 11G MIMO modes for some APs. I've also read the AR71xx platform (the RB400-series CPU) was designed for Atheros's 11N cards, and actually requires kernel hacks to work around bus errors to support the AR5xxx series radio chipsets. If this is true, I suspect switching my busy+crashy RB433H sites over ASAP.

Is it not true that given that purpose of MIMO both antennas should be the same polarity AND same signal strength to assure full throughput? Maybe implementing the signal strenth indicator for each chain WOULD BE a good idea for the next beta release.

mp3turbo · Thu Jul 02, 2009 8:47 am

I second this idea ; I was setting up and aligning both links with standard 802.11a cards and then replacing them with 802.11n just because there is no individual signal strength incidator. At least I didn't find one.

Thu Jul 02, 2009 10:44 am

I second this idea ; I was setting up and aligning both links with standard 802.11a cards and then replacing them with 802.11n just because there is no individual signal strength incidator. At least I didn't find one.

maybe in future releases we will add some signal reading or alignment for each chain, but currently the only option is to select only one chain and align the antenna, then select only the second chain and align, and then enable both chains for full operation.

tokk250 · Thu Jul 02, 2009 3:27 pm

802.11n is very cool

, I test it on 1kilometers connection and the both speed is so good (maximum about 90Mbps UDP traffic in both test), I think, see the picture. My configuration is RB433AH & RB433UAH with R52n cards and 18db dual polarized antena box...

/interface wireless security-profiles
set default authentication-types="" eap-methods=passthrough group-ciphers="" \
    group-key-update=5m interim-update=0s management-protection=disabled \
    management-protection-key="" mode=none name=default \
    radius-eap-accounting=no radius-mac-accounting=no \
    radius-mac-authentication=no radius-mac-caching=disabled \
    radius-mac-format=XX:XX:XX:XX:XX:XX radius-mac-mode=as-username \
    static-algo-0=none static-algo-1=none static-algo-2=none static-algo-3=\
    none static-key-0="" static-key-1="" static-key-2="" static-key-3="" \
    static-sta-private-algo=none static-sta-private-key="" \
    static-transmit-key=key-0 supplicant-identity=MikroTik tls-certificate=\
    none tls-mode=no-certificates unicast-ciphers="" wpa-pre-shared-key="" \
    wpa2-pre-shared-key=""
/interface wireless
set 0 ack-timeout=dynamic adaptive-noise-immunity=none allow-sharedkey=no \
    antenna-gain=20 antenna-mode=ant-b area="" arp=enabled band=5ghz-onlyn \
    basic-rates-a/g="" basic-rates-b="" comment="" compression=no country=\
    "slovak republic" default-ap-tx-limit=0 default-authentication=yes \
    default-client-tx-limit=0 default-forwarding=yes dfs-mode=none \
    disable-running-check=no disabled=no disconnect-timeout=3s \
    frame-lifetime=0 frequency=5220 frequency-mode=manual-txpower hide-ssid=\
    no ht-ampdu-priorities=0,1,2,3,4,5,6,7 ht-amsdu-limit=8192 \
    ht-amsdu-threshold=8192 ht-basic-mcs=mcs-4,mcs-7,mcs-12,mcs-15 \
    ht-extension-channel=above-control ht-guard-interval=any ht-rxchains=\
    0,1,2 ht-supported-mcs=mcs-4,mcs-7,mcs-12,mcs-15 ht-txchains=0,1,2 \
    hw-fragmentation-threshold=disabled hw-protection-mode=none \
    hw-protection-threshold=0 hw-retries=4 mac-address=00:0C:42:3A:D9:2A \
    max-station-count=2007 mode=ap-bridge mtu=1500 name=wlan1 \
    on-fail-retry-time=100ms periodic-calibration=disabled \
    periodic-calibration-interval=60 preamble-mode=both \
    proprietary-extensions=post-2.9.25 radio-name=000C423AD92A rate-set=\
    configured scan-list=default security-profile=default ssid=ptp-nx-tst \
    station-bridge-clone-mac=00:00:00:00:00:00 supported-rates-a/g="" \
    supported-rates-b="" tx-power=10 tx-power-mode=all-rates-fixed \
    update-stats-interval=disabled wds-cost-range=50-150 wds-default-bridge=\
    none wds-default-cost=100 wds-ignore-ssid=no wds-mode=disabled \
    wmm-support=disabled
/interface wireless manual-tx-power-table
set 0 comment="" manual-tx-powers="1Mbps:17,2Mbps:17,5.5Mbps:17,11M\
    bps:17,6Mbps:17,9Mbps:17,12Mbps:17,18Mbps:17,24Mbps:17,36Mbps:17,48Mbps:17\
    ,54Mbps:17,HT20-1:0,HT20-2:0,HT20-3:0,HT20-4:0,HT20-5:0,HT20-6:0,HT20-7:0,\
    HT20-8:0,HT40-1:0,HT40-2:0,HT40-3:0,HT40-4:0,HT40-5:0,HT40-6:0,HT40-7:0,HT\
    40-8:0"
/interface wireless nstreme
set 0 comment="" disable-csma=no enable-nstreme=yes \
    enable-polling=yes framer-limit=3200 framer-policy=none
/interface wireless align
set active-mode=yes audio-max=-20 audio-min=-100 audio-monitor=\
    00:00:00:00:00:00 filter-mac=00:00:00:00:00:00 frame-size=300 \
    frames-per-second=25 receive-all=no ssid-all=no
/interface wireless sniffer
set channel-time=200ms file-limit=10 file-name="" memory-limit=10 \
    multiple-channels=no only-headers=no receive-errors=no streaming-enabled=\
    no streaming-max-rate=0 streaming-server=0.0.0.0
/interface wireless snooper
set channel-time=200ms multiple-channels=yes receive-errors=no

any ideas to do a greatest speed ?

mp3turbo · Fri Jul 03, 2009 10:45 am

seeing that you are in slovak republic, you are using illegal frequency 5220.

contact me privately, I'm also from slovakia

tokk250 · Fri Jul 03, 2009 11:15 am

seeing that you are in slovak republic, you are using illegal frequency 5220.

contact me privately, I'm also from slovakia

sorry guy this frequency is used only for testing, now it's running on other frequency... I know the rules about it then I change it after this screenshot and hour of testing
but thanks for reply

marceloru · Sat Jul 04, 2009 10:20 pm

802.11n have the same media acces like a/b/g (csma/ca)?

have the same limitation at 30-35 clients per AP?

is most efective new N-stream or 802.11n on PTMP with >30 clients?

I appreciate if someone did test experience

Thanks

InoX · Sat Jul 04, 2009 11:57 pm

could everyone to post TCP results also? UDP traffic is quite small than TCP.

xezen · Mon Jul 06, 2009 9:18 am

can people also post information on what anttennas were used with the test as i have tryed 12 diffrent makes of grids,dishes,pannels,cpes and all of them made a diffrence to my results on the tests one stable udp but not tcp

others stable on tcp and not udp

others not stable at all

and others that are stable with udp and tcp total throughput is not as great as all the test that have been posted here so please if you do have results please post with more infomation on hardware used in the test thanks all for all the information posted

srosen · Tue Jul 07, 2009 6:52 am

Has anyone tried using any form of encryption with 802.11n? I'm seeing poor throughput whenever I try to enable any type of encryption.

I have a couple of RB600s with R52N cards that I'm bench testing. Right now, I'm having generally good performance - Using HT Extension Channel - above control, Nstreme not enabled, signal strength around -40, I'm seeing UDP bandwidth test (one way only) around 170Mbps, TCP around 49 Mbps. When I change to disabling the Extension channel (which will suite our application better since it only uses 20 Mhz channel and gives us adequate throughput), I'm seeing 106 Mbps UDP and 48 Mbps TCP. So far so good.

However, whenever I enable any kind of encryption, the performance goes way down the tubes. With WPA2-PSK, TKIP encryption (unicast and group) throughput with Extension channel disabled and no nstreme goes down to 38 Mbps UDP, 26 Mbps TCP. Using WEP 40bit, the results are even worse - TCP throughput is about 15 Mbps, and UDP testing ends up causing the connection to drop.

My test setup is as follows:

RB600-1 (Bandwidth tester) - Ether 3 <---> RB600-2 - Ether 3

RB600-2 Wlan1 (station) <--> RB600-3 (target test machine) - Wlan1 (AP Bridge)

Its not an ideal setup because I don't have a 4th Routerboard to put behind the AP, but I saw similar results testing RB600-2 to RB600-3 only. I believe I understand what's happening that makes the WEP 40 bit UDP test so erratic. While doing the testing, I see the CPU on RB600-3 go to 100%. At that point, the throughput goes to 0, and the connection drops shortly thereafter. If I disable wlan1 on RB600-3, the CPU goes to normal (a couple %), and it'll stay low when I reenable wlan1 (until I start another test).

So, it appears that WEP 40bit is causing some kind of extremely excessive CPU load. However, this isn't what's constraining the WPA2-PSK test. When I perofrm that test, CPU load on both RB600-2 and RB600-3 are around 30% - 40%.

Any ideas?

My config on the RB600-3 AP is (security profile default is WPA2-PSK):

/interface wireless
set 0 ack-timeout=dynamic adaptive-noise-immunity=none allow-sharedkey=no antenna-gain=0 antenna-mode=ant-a \
area="" arp=enabled band=5ghz-onlyn basic-rates-a/g=6Mbps basic-rates-b=1Mbps comment="" compression=no \
country=no_country_set default-ap-tx-limit=0 default-authentication=yes default-client-tx-limit=0 \
default-forwarding=yes dfs-mode=none disable-running-check=no disabled=no disconnect-timeout=3s \
frame-lifetime=0 frequency=5180 frequency-mode=manual-txpower hide-ssid=no ht-ampdu-priorities=\
0,1,2,3,4,5,6,7 ht-amsdu-limit=8192 ht-amsdu-threshold=8192 ht-basic-mcs=\
mcs-0,mcs-1,mcs-2,mcs-3,mcs-4,mcs-5,mcs-6,mcs-7 ht-extension-channel=disabled ht-guard-interval=any \
ht-rxchains=0,1 ht-supported-mcs=\
mcs-0,mcs-1,mcs-2,mcs-3,mcs-4,mcs-5,mcs-6,mcs-7,mcs-8,mcs-9,mcs-10,mcs-11,mcs-12,mcs-13,mcs-14,mcs-15 \
ht-txchains=0,1 hw-fragmentation-threshold=disabled hw-protection-mode=none hw-protection-threshold=0 \
hw-retries=4 mac-address=00:0C:42:3A:D5:B6 max-station-count=2007 mode=ap-bridge mtu=1500 name=wlan1 \
on-fail-retry-time=100ms periodic-calibration=disabled periodic-calibration-interval=60 preamble-mode=short \
proprietary-extensions=post-2.9.25 radio-name=LJR-Tank-5.8 rate-set=default scan-list=default \
security-profile=default ssid=LJR.WR-1-5.8 station-bridge-clone-mac=00:00:00:00:00:00 supported-rates-a/g=\
6Mbps,9Mbps,12Mbps,18Mbps,24Mbps,36Mbps,48Mbps,54Mbps supported-rates-b=1Mbps,2Mbps,5.5Mbps,11Mbps \
tx-power-mode=default update-stats-interval=disabled wds-cost-range=50-150 wds-default-bridge=none \
wds-default-cost=100 wds-ignore-ssid=no wds-mode=disabled wmm-support=disabled
/interface wireless manual-tx-power-table
set wlan1 comment="" manual-tx-powers="1Mbps:17,2Mbps:17,5.5Mbps:17,11Mbps:17,6Mbps:17,9Mbps:17,12Mbps:17,18Mbps:\
17,24Mbps:17,36Mbps:17,48Mbps:17,54Mbps:17,HT20-1:0,HT20-2:0,HT20-3:0,HT20-4:0,HT20-5:0,HT20-6:0,HT20-7:0,HT2\
0-8:0,HT40-1:0,HT40-2:0,HT40-3:0,HT40-4:0,HT40-5:0,HT40-6:0,HT40-7:0,HT40-8:0"
/interface wireless nstreme
set wlan1 comment="" disable-csma=no enable-nstreme=no enable-polling=yes framer-limit=3200 framer-policy=none

The client is configured similarly.

diogomaia · Fri Jul 10, 2009 7:47 pm

on 10 km on 433 ah 800 mhz, sr 71 a