Page 1 of 1

Significant improvement for 60 GHz solutions

Posted: Wed Mar 21, 2018 3:49 pm
by strods
We have made a significant improvement for 60 GHz wireless (Wireless Wire and wAP 60G) configurations in the latest 6.42rc version release - increased supported distance for wAP 60G to 200+ meters.

Those of you who are willing to test rc versions on your test network, please upgrade to the latest version, test the performance and report to support@mikrotik.com with the results. Your effort and feedback will be highly appreciated.

Please note that rc versions are released strictly for testing purposes and should not be used on crucial network devices.

Re: Significant improvement for 60 GHz solutions

Posted: Wed Mar 21, 2018 5:21 pm
by blingblouw
Could you guys be a bit more specific as to what we are looking for?

Do you mean you can maintain lower mcs rates at higher distance or are you saying we should get full gig at 200+ meters?

What exactly are we looking for here? Also, ptp or ptmp?

Re: Significant improvement for 60 GHz solutions

Posted: Wed Mar 21, 2018 9:25 pm
by djvolt1942
I join the question, what exactly are we looking for? Stability at distances +200m on wAPs or a better signal?

Re: Significant improvement for 60 GHz solutions

Posted: Thu Mar 22, 2018 9:49 am
by Bergante
I have tried with my 20 - 30 m short link.

With 6.43rc49 I see it hitting MCS 10. With the previous versions the maximum it reached was MCS8.

Re: Significant improvement for 60 GHz solutions

Posted: Thu Mar 22, 2018 1:39 pm
by normis
In simple words, you can get ~900Mbits not only at 100 meters, but now even at 200 meters.

Re: Significant improvement for 60 GHz solutions

Posted: Thu Mar 22, 2018 1:48 pm
by Bergante
In simple words, you can get ~900Mbits not only at 100 meters, but now even at 200 meters.
But we are curious souls who would like to Know some details! ;)

Now seriously, I have spammed support with some tests I have done in my link, which is a short one. Stability has gone down with rc49. With rc30 it was rock
solid, as it is with 6.41.3.

The changes I noticed are, higher TX power? (Assuming tx-sector is an indication of TX power)

Re: Significant improvement for 60 GHz solutions

Posted: Thu Mar 22, 2018 5:57 pm
by human1982
350m how can i get ?

Re: Significant improvement for 60 GHz solutions

Posted: Thu Mar 22, 2018 6:51 pm
by MonkeyDan
I installed rc49 on a few of our >100m links (PtP and PtMP) and neither MCS or Signal Quality have changed. Throughput seems better, though, as running a traffic generator over our longest link (160m?) nets a solid 500Mbps full-duplex with little fluctuation. Used to be all over the place. Seeing a lot more link downs, too, but that might be due to rain we've been having. Should we be seeing better MCS and/or Quality readings? What exactly is being improved?

Re: Significant improvement for 60 GHz solutions

Posted: Wed Apr 04, 2018 12:26 am
by antonsb
I installed rc49 on a few of our >100m links (PtP and PtMP) and neither MCS or Signal Quality have changed. Throughput seems better, though, as running a traffic generator over our longest link (160m?) nets a solid 500Mbps full-duplex with little fluctuation. Used to be all over the place. Seeing a lot more link downs, too, but that might be due to rain we've been having. Should we be seeing better MCS and/or Quality readings? What exactly is being improved?
Calibration data for beamforming patterns.

Re: Significant improvement for 60 GHz solutions

Posted: Mon Apr 09, 2018 10:21 pm
by MonkeyDan
Just a heads up, I had an issue with 6.42rc56 on a multipoint (ap-bridge) AP. Interfaces weren't showing up and the station wasn't connecting. I switched to bridge mode and everything worked again. Really digging the rssi and tx-sector-info, but I don't think it's ready for ap-bridge mode yet.

Re: Significant improvement for 60 GHz solutions

Posted: Tue Apr 10, 2018 8:48 am
by Lupin
The perfect product for the 60Ghz, would be a point-to-multipoint solution, clearly with high gain CPE antennas for distances up to max 1km.
Possibility to connect up to 50/60 customers on a single cell
Channels smaller than the 2000Mhz (1000Mhz or 500Mhz)
TxPower increased to 20dB
A integrated failover 5ghz radio could be failover when primary 60Ghz fail due to rain

With this solution provider wisp can delivery 100/200/500 Mbps to customers and fight vdsl fttc/ftth of cable isp

Re: Significant improvement for 60 GHz solutions

Posted: Tue Apr 10, 2018 9:36 am
by server8
Lupin you are right I think we need a solutions like ignitenet with 5 ghz backup to have a killer product to use in PtMP scenario to max 700m I hope to see it soon it's not an hard job make it

Re: Significant improvement for 60 GHz solutions

Posted: Tue Apr 10, 2018 9:42 am
by Bergante
Just a heads up, I had an issue with 6.42rc56 on a multipoint (ap-bridge) AP. Interfaces weren't showing up and the station wasn't connecting. I switched to bridge mode and everything worked again. Really digging the rssi and tx-sector-info, but I don't think it's ready for ap-bridge mode yet.
Confirmed. The interface hangs if I set the mode to ap-bridge.

Re: Significant improvement for 60 GHz solutions

Posted: Tue Apr 10, 2018 9:49 am
by pospanko
The perfect product for the 60Ghz, would be a point-to-multipoint solution, clearly with high gain CPE antennas for distances up to max 1km.
Possibility to connect up to 50/60 customers on a single cell
Channels smaller than the 2000Mhz (1000Mhz or 500Mhz)
TxPower increased to 20dB
A integrated failover 5ghz radio could be failover when primary 60Ghz fail due to rain

With this solution provider wisp can delivery 100/200/500 Mbps to customers and fight vdsl fttc/ftth of cable isp
Totally agreed.

Re: Significant improvement for 60 GHz solutions

Posted: Tue Apr 10, 2018 3:13 pm
by antonsb
Just a heads up, I had an issue with 6.42rc56 on a multipoint (ap-bridge) AP. Interfaces weren't showing up and the station wasn't connecting. I switched to bridge mode and everything worked again. Really digging the rssi and tx-sector-info, but I don't think it's ready for ap-bridge mode yet.
We are currently working on fixing this, fixes will be included in upcoming versions.

Re: Significant improvement for 60 GHz solutions

Posted: Wed Apr 11, 2018 5:45 pm
by mistry7
The perfect product for the 60Ghz, would be a point-to-multipoint solution, clearly with high gain CPE antennas for distances up to max 1km.
Possibility to connect up to 50/60 customers on a single cell
Channels smaller than the 2000Mhz (1000Mhz or 500Mhz)
TxPower increased to 20dB
A integrated failover 5ghz radio could be failover when primary 60Ghz fail due to rain

With this solution provider wisp can delivery 100/200/500 Mbps to customers and fight vdsl fttc/ftth of cable isp
I hoped to see something like that at mum, but nothing

Re: Significant improvement for 60 GHz solutions

Posted: Wed Apr 11, 2018 8:09 pm
by gr0mit
I'm kinda keen to see some more helpful SNMP OIDs soon - as it would be really useful to see how the link behaves in bad weather. e.g. coding scheme, and RX and Tx sig strength?

Re: Significant improvement for 60 GHz solutions

Posted: Thu Apr 12, 2018 6:22 pm
by wpeople
I would be pleased to have on option in the Wireless60G to
- run a script (like netwatch)
- or disable/enable the bridge port status depending on Signal (or CCQ in future)

This would be useful to switch to backup before link quality goes down to loosing packets.

Re: Significant improvement for 60 GHz solutions

Posted: Thu Apr 12, 2018 6:51 pm
by npero
You can do it that now just run script in scheduler read RSSI or signal and disable or enable port in bridge or whatever you wont it is possible now with script run in scheduler.

Re: Significant improvement for 60 GHz solutions

Posted: Thu Apr 12, 2018 10:59 pm
by wpeople
yes, i know, i can run a scheduler.
however generally in wireless would be a nice option to offer a (netwatch alike) option to run script in case of
- signal (or ccq) is decreasing and reached X value (disable in bridge or dyn.routing)
- signal (or ccq) is increasing and reached Y value (enable in brdige or dyn.routing)

Would be nice even with general wireless package.

Re: Significant improvement for 60 GHz solutions

Posted: Fri Apr 13, 2018 11:08 am
by Bergante
I'm kinda keen to see some more helpful SNMP OIDs soon - as it would be really useful to see how the link behaves in bad weather. e.g. coding scheme, and RX and Tx sig strength?
The coding scheme is available through : MCS

Look at the published MIB variables: https://wiki.mikrotik.com/wiki/Manual:I ... monitoring

Re: Significant improvement for 60 GHz solutions

Posted: Fri Apr 20, 2018 3:01 pm
by vladimirslk
sofar I am happy with 60ghz dishes, hopefully there would not be any bugs with them to run on production : )
still under big question, will they beat 2.7-3km range or not, so far as I calculated should be on the limits, with fail-over to existing 5ghz could be a good improvement.

Re: Significant improvement for 60 GHz solutions

Posted: Tue Apr 24, 2018 3:42 pm
by mlenhart
increased supported distance for wAP 60G to 200+ meters.
@strods: does this apply to all units, since the beginning of product distributions or is it limited only to some HW version? I know it has been reached by SW modification and phase array patterns, just want to be sure it is not HW dependent.

Thank you.

Re: Significant improvement for 60 GHz solutions

Posted: Tue Apr 24, 2018 4:33 pm
by macgaiver
@strods: does this apply to all units, since the beginning of product distributions or is it limited only to some HW version? I know it has been reached by SW modification and phase array patterns, just want to be sure it is not HW dependent.

Thank you.
Topic is clearly about software improvements

Re: Significant improvement for 60 GHz solutions

Posted: Tue Apr 24, 2018 10:19 pm
by mlenhart
Topic is clearly about software improvements
It is OK, that it is about SW improvements, but this SW improvements improves HW specifications. Maybe I was not clear enough, so once again (hope you will not consider it as spamming...)
Does the improvement for 200 meters apply for all WAP60G units which have been produced (regardless of license level) or is it valid only for some HW revisions (if there are any...) of WAP60G?

Thank you.

Re: Significant improvement for 60 GHz solutions

Posted: Tue Apr 24, 2018 11:46 pm
by mkx
Improvements in SW many times improve overall functionality. Therefore one can not say what's HW specification... better to refer to it as "system" specification.

In my company (I work for mobile division of incumbent telco) we have an internal joke about vendors: "yet again they managed to speed-up execution of infinite loop". And that particular (world-class) vendor officially changes capacity figures in their documentation ... or better yet, the specification explicitly states the SW version it's valid for.

So yes, SW improvements can considerably improve performance and capacity without any HW change.

Re: Significant improvement for 60 GHz solutions

Posted: Wed Apr 25, 2018 9:17 am
by Bergante
Improvements in SW many times improve overall functionality. Therefore one can not say what's HW specification... better to refer to it as "system" specification.
(...)
So yes, SW improvements can considerably improve performance and capacity without any HW change.
Remember that a phased array is the antenna rough equivalent of a FPGA chip. So, just reconfiguring it can be equivalent to installing a new antenna.

In my case (units purchased in February) the change was so dramatic that it caused overload problems (my link is short, 20 - 30 m across a street). I had to reduce power, go figure! :)

Re: Significant improvement for 60 GHz solutions

Posted: Fri Apr 27, 2018 8:33 am
by komenski
For all WAP60g users please Mikrotik to explain the TX-sector numbers. For use of PtMP or colocation, it would be very useful to know the angles and directions of radiation. Please also add the option to modify the RX sector or turn off the beamforming routine. Thanks

Re: Significant improvement for 60 GHz solutions

Posted: Fri Apr 27, 2018 5:20 pm
by antonsb
For all WAP60g users please Mikrotik to explain the TX-sector numbers. For use of PtMP or colocation, it would be very useful to know the angles and directions of radiation. Please also add the option to modify the RX sector or turn off the beamforming routine. Thanks
Explanation is already available in monitor:
for example:
 
 tx-sector: 29
 tx-sector-info: 
 right 0.6 degrees, down 0.2 degrees

Re: Significant improvement for 60 GHz solutions

Posted: Sun Apr 29, 2018 9:42 pm
by komenski
So, tx-sector=29 means radiations angle 29 degrees, really??

Re: Significant improvement for 60 GHz solutions

Posted: Sun Apr 29, 2018 10:37 pm
by r00t
No, tx-sector 29 means +0.6 degrees azimuth, -0.2 degrees elevation. tx-sector is just a sector (beam configuration) identifier that translates to some predefined azimuth/elevation combination, as output to the console.

Re: Significant improvement for 60 GHz solutions

Posted: Wed May 02, 2018 2:48 am
by wispwest
I have a pair of WAP-60G and have modified some reflectors to increase range up to 1-mile, is there still any distance LIMIT in the software that would prevent them from even registering if they could see each other at 1600 meters? Do I need the latest RC firmware?

Re: Significant improvement for 60 GHz solutions

Posted: Mon May 28, 2018 10:36 am
by normis
I have a pair of WAP-60G and have modified some reflectors to increase range up to 1-mile, is there still any distance LIMIT in the software that would prevent them from even registering if they could see each other at 1600 meters? Do I need the latest RC firmware?
As said a few times in other topics, this will not work well. The devices are precisely calibrated so that the beamforming antenna array knows exactly how to adjust itself. If you put that into a reflector, it gets really confused. It may work sometimes, but most likely it will not work.

Re: Significant improvement for 60 GHz solutions

Posted: Mon May 28, 2018 11:09 am
by Bergante
For a reflector to work properly you really need to "flood" it with radiation.

If the antenna main lobe is more like a sharp pencil the reflector won't be so effective.

https://en.wikipedia.org/wiki/Parabolic ... ed_pattern

Re: Significant improvement for 60 GHz solutions

Posted: Mon May 28, 2018 5:57 pm
by stephanhd
here my data (6.43rc19) - hope this helps; I am willing to perform other tests...

frequency: 58320
remote-address: 30:07:xx
mcs: 8
phy-rate: 2.3Gbps
signal: 95
rssi: -51
tx-sector: 27
tx-sector-info: right 3.8 degrees, down 3.8 degrees
rx-sector: 96
distance: 15.12m (it's an "over the street" scenario, having not 100% clear sight)

regards, Stephan

Re: Significant improvement for 60 GHz solutions

Posted: Sun Jun 10, 2018 12:08 pm
by anuser
As I just read about new chipsets for Smartphones with 802.11ad Support.

Think about a new room with 200 students.
- In the ceiling there will 9 double sockets with RJ45
- I'll add 9x cAP AC to those with lowered sending power...
- I will also have some RJ45 sockets.

Ist WAP 60G meant to work like cAP AC with CAPSMAN enabled, multiple concurrent SSIDs enabled and support for 20/30 smartphones connected?

Re: Significant improvement for 60 GHz solutions

Posted: Mon Jun 11, 2018 12:28 pm
by antonsb
As I just read about new chipsets for Smartphones with 802.11ad Support.

Think about a new room with 200 students.
- In the ceiling there will 9 double sockets with RJ45
- I'll add 9x cAP AC to those with lowered sending power...
- I will also have some RJ45 sockets.

Ist WAP 60G meant to work like cAP AC with CAPSMAN enabled, multiple concurrent SSIDs enabled and support for 20/30 smartphones connected?
currently other vendors are not supported, as well as CAPsMAN.
This will be added in future releases of RouterOS

Re: Significant improvement for 60 GHz solutions

Posted: Mon Jun 25, 2018 4:34 pm
by cliffordjames
Improvements in SW many times improve overall functionality. Therefore one can not say what's HW specification... better to refer to it as "system" specification.
(...)
So yes, SW improvements can considerably improve performance and capacity without any HW change.
Remember that a phased array is the antenna rough equivalent of a FPGA chip. So, just reconfiguring it can be equivalent to installing a new antenna.

In my case (units purchased in February) the change was so dramatic that it caused overload problems (my link is short, 20 - 30 m across a street). I had to reduce power, go figure! :)
How did you reduce power?

Re: Significant improvement for 60 GHz solutions

Posted: Tue Jun 26, 2018 10:05 am
by Bergante
How did you reduce power?

/interface w60g set 0 tx-power=[1..15] or default.

In my case set 0 tx-power=10 works great, it links fine with 5 as well.

However, be careful with this setting when updating firmware. If they make changes to the antenna patterns you may need to change/remove it.

Re: Significant improvement for 60 GHz solutions

Posted: Mon Jul 16, 2018 7:16 pm
by komenski
I have a pair of WAP-60G and have modified some reflectors to increase range up to 1-mile, is there still any distance LIMIT in the software that would prevent them from even registering if they could see each other at 1600 meters? Do I need the latest RC firmware?
As said a few times in other topics, this will not work well. The devices are precisely calibrated so that the beamforming antenna array knows exactly how to adjust itself. If you put that into a reflector, it gets really confused. It may work sometimes, but most likely it will not work.
The correct answer should sound, Mikrotik does not want it, unfortunately. There is no technical reason why the WAP60g does not work well in the focus of the parabola. When receiving (RX) everything is OK, unfortunately the transmission angle (TX) is set to a much smaller angle, so the parabols are not very bright. If it was possible to set the beam angle (TX) fixed or much larger, it would be after the problem. Well, maybe we'll get a special version of FW, WAP60g in the focus of the parabola ...

Re: Significant improvement for 60 GHz solutions

Posted: Tue Jul 17, 2018 10:52 pm
by buzted
RBLHGG-60ad 6.43rc44

650 Meters 642Mbps/474Mbps

admin@MikroTik] /interface w60g> monitor 0
connected: yes
frequency: 58320
remote-address: 04:D6:AA:AF:D3:7E
mcs: 8
phy-rate: 2.3Gbps
signal: 95
rssi: -52
tx-sector: 29
tx-sector-info: right 0.6 degrees, down 0.2 degrees
-- [Q quit|D dump|C-z pause]

admin@MikroTik] /interface w60g> monitor 0
connected: yes
frequency: 58320
remote-address: 04:D6:AA:AF:D2:95
mcs: 8
phy-rate: 2.3Gbps
signal: 95
rssi: -50
tx-sector: 16
tx-sector-info: left 1.4 degrees, down 0.6 degrees
-- [Q quit|D dump|C-z pause]

Re: Significant improvement for 60 GHz solutions

Posted: Thu Jul 19, 2018 7:51 pm
by antonsb
results
Test with external tool or Traffic-generator and You will get full Wire speed

Re: Significant improvement for 60 GHz solutions

Posted: Mon Oct 01, 2018 9:26 pm
by Michaelcrapse
results
Test with external tool or Traffic-generator and You will get full Wire speed
I don't. 6.43rc64.
Real world TCP/IP throughput is only 200-300mbps down, 200-300 up, tested using Iperf3 only over the link, and speedtest.net to 700mbps fed pop. Using udp test get around 600-700 download one way(across the link only)

Re: Significant improvement for 60 GHz solutions

Posted: Wed Jan 23, 2019 5:33 pm
by anuser
As I just read about new chipsets for Smartphones with 802.11ad Support.

Think about a new room with 200 students.
- In the ceiling there will 9 double sockets with RJ45
- I'll add 9x cAP AC to those with lowered sending power...
- I will also have some RJ45 sockets.

Ist WAP 60G meant to work like cAP AC with CAPSMAN enabled, multiple concurrent SSIDs enabled and support for 20/30 smartphones connected?
currently other vendors are not supported, as well as CAPsMAN.
This will be added in future releases of RouterOS
For now I´m planing with 802.11ad for that mentioned room. Do you plan this feature to be released in 2019? Would this be realistic?

Re: Significant improvement for 60 GHz solutions

Posted: Thu Feb 07, 2019 10:24 am
by antonsb
please show us exactly how are you testing this. We get full speed on much higher distances by using same testing tools.
make sure that testing PCs has enough power, if possible run test in multi core mode, use multiple streams as well.

Re: Significant improvement for 60 GHz solutions

Posted: Sat Sep 21, 2019 3:32 pm
by nel3
In simple words, you can get ~900Mbits not only at 100 meters, but now even at 200 meters.
Hi Normis and friends,

Question: 1
Is this speed ~900Mbps only single upload or download bandwidth tests or both directions?

Sinario: 1 I need advice on please?
I setup a 60GHz point to point link, with RB4011iGS+ connected to the links on either side. (Routers not in a bridge configuration IP's directly on the interface of router port)

Equipment:
CAT6 shielded outdoor cable. (3 meters) both high site and CPE simulation sides.
Shielded CAT6 RJ45 connectors.
24 volt power feed.

Distance between links was 160 meters signal was between 90 and 100.
TCP test: Freq 58320
Upload 940Mbps
Download 940Mbps
Both directions 884 / 885 Mbps

0 packet loss

I did the same TCP both directions (Upload and Download) test directly between the 2 routers with a half meter CAT5e cable and the speeds are 918 / 917 Mbps

Is this the fastest throughput i can expect through the links my boss would like to know?

I am happy with the testing results.

Is there any other settings that I can adjust besides ensuring good alignment and change between the four frequencies available, with router OS 6.44.5?

Thank you in advance!