Community discussions

 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

As we would like to evolve the wireless

Wed Sep 04, 2013 12:23 pm

Requests that we would like to be implemented in RouterOS (software and hardware) ONLY for Wireless section

SOFTWARE:
1 - Scan to a File: save on the file the results obtained with timed scan instruments like "Scan", "Freq.Usage", "Snooper", "Spectral Scan", "Spectral History" (disconnecting itself) and then log back in and see/download the results ... Status: PARTIAL WORK IN PROGRESS!
UPDATE 29/04/2014: Spectral Scan allow to save a file on the RB but the are working for a viewer to see this proprietary file http://forum.mikrotik.com/viewtopic.php ... 36#p423336
2 - Detailed logs: When the radio disconnected (for example "Extensive Data Loss") in the log we need to see the last signal Tx / Rx registered before they disconnected
3 - Improved NV2 latency: Lower latency of NV2 protocol when not in use. With a normal ping (without traffic on the radio) you can find spike and jitter is unstable
4 - ATPC: it often happens that the radio, when installed close to the AP, makes poor the performance of the AP that received signal too strong. We would need to be able to set a maximum signal threshold (for example,-55dB) to make sure that the radio (the SXT for example) could auto adjust the output power until it reaches that threshold (if possible the autoregulation must not disconnect the link)
5 - GPS sync: possibility to reuse the same frequencies without interference (in mode: A+B+A+B)
6 - Support the new protocol 802.11ac with channels from 20MHz (and 40Mhz) up to 256QAM ... Status: COMPLETE!
UPDATE 20/08/2014: Protocol is ready to use
7 - 10/20/40mhz mode in channel-width parameter, in addition to classic 5MHz, 10MHz, 20MHz; 20/40mhz-ht-above; 20/40mhz-ht-below, or also the 10/20mhz. In certain situations "10Mhz" performs better than the 20Mhz: they have more stability (without using the Wireless> Channels)
8 - Having the graphs of the historical signal-strength to see
9 - Scan starting point: Have client/remote start scanning on last used frequency or simply in the the frequency setting in the client/remote as a starting point. This would speed up reconnects a lot particularly with 5ghz where you may have a long scan list. (by ejansson)
10 - Improve NV2 throughput: for example the motorola canopy pmp450 that allow an aggregate traffic up to 90Mbps in ptmp (by ejansson)
11 - NV2 manual Airtime: would like to slow down cpes which have bad signal to save airtime for good installed cpes (by ste)
12 - Advanced Scanning: Scans all visible proprietary APs and for each (respecting SSID, security settings, connect lists, etc.) connects and keeps track of the SNR, Rx/Tx signal strength, radioName, mac-address and distance.
In a normal scan the TX Signal Strenght and distance is not visible and the results show also all other operators APs
13 - Wireless Controller feature allows to centralize wireless network management and data processing. See http://goo.gl/wwYFqX ... Status: COMPLETE!
UPDATE 29/04/2014: Add the CAPsMAN feature http://wiki.mikrotik.com/wiki/Manual:CAPsMAN
14 - Auto frequency feature added by mikrotik that select the best channel to use ... Status: COMPLETE!

HARDWARE:
1 - The routerboard http://routerboard.com/RB912UAG-5HPnD-OUT is very nice but it would need a shield one, with the same case (only in metal) in order to decrease the co-location self-interference ... Status: COMPLETE!
UPDATE 20/08/2014: NetMetal is coming out
2 - A dual-pol cpe (like Sexant) 23dB panel like this http://www.cyberteam.pl/en/produkt/show ... Dual_/MMCX , but with the ability to see the LEDs array for alignment use ... Status: COMPLETE!
UPDATE 29/04/2014: QRT-5 is ready for use http://routerboard.com/RB911G-5HPnD-QRT If possibile we need a low cost (Level 3 + NO_Gigabit version for customer's use)
3 - 2+5Ghz CPE: for example a Sexant with a new model of RB411R (integrated 2Ghz with 802.11n protocol) and a slot with a 5Ghz radio. External N connector for a 2Ghz antenna (by haik01)


If you have other ideas or suggestions we could keep this post as sticky

Regards
Mirko
Last edited by Lupin on Wed Aug 20, 2014 10:47 pm, edited 8 times in total.
 
Muqatil
Trainer
Trainer
Posts: 574
Joined: Mon Mar 03, 2008 1:03 pm
Location: London - UK
Contact:

Re: As we would like to evolve the wireless

Wed Sep 04, 2013 1:14 pm

+1 on all of em.
It feels like the wireless part of MikroTik fell apart lately...
Renato Bernardi

skype: medtech5
 
ejansson
Member
Member
Posts: 301
Joined: Fri Oct 21, 2005 4:09 pm
Location: Manitoba, Canada

Re: As we would like to evolve the wireless

Wed Sep 04, 2013 4:49 pm

couldn't agree more on all your points
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Sat Sep 07, 2013 1:27 pm

If you have other ideas or improvements to report, write!! that I keep updated the list above
8)
 
0ldman
Forum Guru
Forum Guru
Posts: 1446
Joined: Thu Jul 27, 2006 5:01 am

Re: As we would like to evolve the wireless

Sat Sep 07, 2013 4:58 pm

All good ideas.

I haven't tried it yet, but I was thinking about using WDS to sync as a work around.
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Sat Sep 07, 2013 6:52 pm

All good ideas.

I haven't tried it yet, but I was thinking about using WDS to sync as a work around.
Exactly, what you mean by wds sync?
 
ejansson
Member
Member
Posts: 301
Joined: Fri Oct 21, 2005 4:09 pm
Location: Manitoba, Canada

Re: As we would like to evolve the wireless

Mon Sep 09, 2013 8:22 pm

1) Have client/remote start scanning on last used frequency or simply in the the frequency setting in the client/remote as a starting point. This would speed up reconnects a lot particularly with 5ghz where you may have a long scan list.

2) Ability to choose to have the client/remote to change and scan other channels size if it can connect back to the ap or master. This would reduce the work needed for major changes that some times need to be done.

3) Improve drivers/software for NV2. Right now stability is much better than Nstreme but through put can be half of nstreme too with P2MP. Just link to see more effort put in here in terms of optimization.

4) Sync; while tower to tower would be nice local is a much simpler thing to do and would go a long way to improving things at congested sites.

Hardware:

1) Ethernet surge protection, the weakest part of any RB product. Lost lots of ports on boards over the years. built in protection on the board is cheap and makes installs faster and easier.

2) Better filtering on RF.

3) Consider a "premium line" for serious wisps that don't what the cheapest thing but a top quality product at a reasonable price.
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Tue Sep 10, 2013 12:35 am

I am happy that someone posts his ideas. thanks ejansson. :D
We try to understand your points:
1) Have client/remote start scanning on last used frequency or simply in the the frequency setting in the client/remote as a starting point. This would speed up reconnects a lot particularly with 5ghz where you may have a long scan list.
This is a good optimization
2) Ability to choose to have the client/remote to change and scan other channels size if it can connect back to the ap or master. This would reduce the work needed for major changes that some times need to be done.
This is similar to my point number 7
3) Improve drivers/software for NV2. Right now stability is much better than Nstreme but through put can be half of nstreme too with P2MP. Just link to see more effort put in here in terms of optimization.
We could say "improve the throughput". For example the motorola canopy pmp450 that allow an aggregate traffic up to 90Mbps in pmp
4) Sync; while tower to tower would be nice local is a much simpler thing to do and would go a long way to improving things at congested sites.
You mean a synchronization via Ethernet or tcp/ip?
1) Ethernet surge protection, the weakest part of any RB product. Lost lots of ports on boards over the years. built in protection on the board is cheap and makes installs faster and easier.
This is not a characteristic related to wireless
2) Better filtering on RF.
Exactly what you mean?
3) Consider a "premium line" for serious wisps that don't what the cheapest thing but a top quality product at a reasonable price.
You mean a product built with more durable components? (Give an example for wireless section)
 
dzieva
Frequent Visitor
Frequent Visitor
Posts: 51
Joined: Fri Mar 25, 2011 4:51 pm
Location: Iguassu Falls, Brasil

Re: As we would like to evolve the wireless

Fri Sep 13, 2013 2:36 am

Improved NV2 now...
 
User avatar
Bergante
Member Candidate
Member Candidate
Posts: 131
Joined: Tue Feb 28, 2012 12:27 pm
Location: Bilbao, Spain

Re: As we would like to evolve the wireless

Fri Sep 13, 2013 2:21 pm

A very important feature would be programmable hardware attenuation for the receiver part.

That would improve the performance in high interference places. You would be able to use a higher transmission power, so that your signal would stay above the interference, at the same time avoiding the receiver overload that could happen.

I mean, improving S/N ratio without the risk of overloading the receiver. A stepped attenuator with a maximum of 20 dB would be great I think.

Maybe some wireless chips include it?
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Fri Sep 13, 2013 10:38 pm

You mean an attenuator on the 5Ghz frequency (the ones where you work normally) or attenuation larger or smaller 5Ghz frequency (or the specified range)

I believe that an attenuator lowers the input power and also output when used on the range of operating frequencies... you give an example
 
haik01
Member
Member
Posts: 406
Joined: Sat Mar 23, 2013 10:25 am
Location: Netherlands

Re: As we would like to evolve the wireless

Sat Sep 14, 2013 12:12 pm

A nice one, is a routerboard, with integrated 2,4 and 5 Ghz in the same box.
 
User avatar
honzam
Forum Guru
Forum Guru
Posts: 2293
Joined: Wed Feb 27, 2008 10:27 pm
Location: Czech Republic

Re: As we would like to evolve the wireless

Sat Sep 14, 2013 12:18 pm

Requests that we would like to be implemented in RouterOS (software and hardware) ONLY for Wireless section

SOFTWARE:
1 - Scan to a File: save on the file the results obtained with timed scan instruments like "Scan", "Freq.Usage", "Snooper", "Spectral Scan", "Spectral History" (disconnecting itself) and then log back in and see/download the results.
very useful +1
2 - Detailed logs: When the radio disconnected (for example "Extensive Data Loss") in the log we need to see the last signal Tx / Rx registered before they disconnected
AND RADIO NAME
3 - Improved NV2 latency: Lower latency of NV2 protocol when not in use. With a normal ping (without traffic on the radio) you can find spike and jitter is unstable
+ 1 we need better latency
6 - Support the new protocol 802.11ac with channels from 20MHz (and 40Mhz) up to 256QAM
+1 we still wait
LAN, FTTx, Wireless. ISP operator
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Sat Sep 14, 2013 10:43 pm

I update the first post with some of your suggestions 8)

If I have not understood well, correct me
 
User avatar
Bergante
Member Candidate
Member Candidate
Posts: 131
Joined: Tue Feb 28, 2012 12:27 pm
Location: Bilbao, Spain

Re: As we would like to evolve the wireless

Mon Sep 16, 2013 11:50 am

You mean an attenuator on the 5Ghz frequency (the ones where you work normally) or attenuation larger or smaller 5Ghz frequency (or the specified range)

I believe that an attenuator lowers the input power and also output when used on the range of operating frequencies... you give an example
An attenuator on the frequency band where you are working, and affecting *only* the receive section.

An example: You are setting up a wireless link on a noisy place. You crank up the transmitter power (either by increasing power or using a more directional antenna) but still you don't have a good enough S/N ratio. If the received signal is too strong, the receiver will overload.

In that case, if you have, for example, a 10 dB attenuator on the receiver, you have that extra margin before overload. You enable the attenuator and it will attenuate both the surronding interference and the interesting signal. But if you add gain and/or power to the interesting signal, you are effectively increasing the S/N ratio, which is what you need to achieve. The attenuator gives you a greater headroom for that.
 
ste
Forum Guru
Forum Guru
Posts: 1813
Joined: Sun Feb 13, 2005 11:21 pm

Re: As we would like to evolve the wireless

Mon Sep 16, 2013 12:57 pm

My list is nearly the same ;-))
Requests that we would like to be implemented in RouterOS (software and hardware) ONLY for Wireless section

SOFTWARE:
1 - Scan to a File: save on the file the results obtained with timed scan instruments like "Scan", "Freq.Usage", "Snooper", "Spectral Scan", "Spectral History" (disconnecting itself) and then log back in and see/download the results.
2 - Detailed logs: When the radio disconnected (for example "Extensive Data Loss") in the log we need to see the last signal Tx / Rx registered before they disconnected
3 - Improved NV2 latency: Lower latency of NV2 protocol when not in use. With a normal ping (without traffic on the radio) you can find spike and jitter is unstable
I would also like to see some sort of priorization in nv2. I would like to slow down cpes which have bad signal to save airtime for good installed cpes.
4 - ATPC: it often happens that the radio, when installed close to the AP, makes poor the performance of the AP that received signal too strong. We would need to be able to set a maximum signal threshold (for example,-55dB) to make sure that the radio (the SXT for example) could auto adjust the output power until it reaches that threshold (if possible the autoregulation must not disconnect the link)
I've a script for this but changing the power/antenna gain cause a reconnect.

5 - GPS sync: possibility to reuse the same frequencies without interference (in mode: A+B+A+B)
6 - Support the new protocol 802.11ac with channels from 20MHz (and 40Mhz) up to 256QAM
MuMIMO in 802.11ac would be a great feature.

7 - 10/20/40mhz mode in channel-width parameter, in addition to classic 5MHz, 10MHz, 20MHz; 20/40mhz-ht-above; 20/40mhz-ht-below, or also the 10/20mhz. In certain situations "10Mhz" performs better than the 20Mhz: they have more stability (without using the Wireless> Channels)
8 - Having the graphs of the historical signal-strength to see
9 - Scan starting point: Have client/remote start scanning on last used frequency or simply in the the frequency setting in the client/remote as a starting point. This would speed up reconnects a lot particularly with 5ghz where you may have a long scan list. (by ejansson)
10 - Improve NV2 throughput: for example the motorola canopy pmp450 that allow an aggregate traffic up to 90Mbps in ptmp (by ejansson)
We should be able to see 90MBps in 10MHz Channel and 180MBPs in 20MHz with *optimal* conditions.
(pmp450 are very limited with SM in NATmode, guess it's a weak CPU).

HARDWARE:
1 - The routerboard http://routerboard.com/RB912UAG-5HPnD-OUT is very nice but it would need a shield one, with the same case (only in metal) in order to decrease the co-location self-interference
We prefer solutions without external Coax-Cables. So integrated Antenna with box on the
back connected with short pigtails within the housing.
The new SXT Sector with a better housing/antenna would be a good solution.
2 - A dual-pol cpe (like Sexant) 23dB panel like this http://www.cyberteam.pl/en/produkt/show ... Dual_/MMCX , but with the ability to see the LEDs array for alignment use.
3 - 2+5Ghz CPE: for example a Sexant with a new model of RB411R (integrated 2Ghz with 802.11n protocol) and a slot with a 5Ghz radio. External N connector for a 2Ghz antenna (by haik01)


If you have other ideas or suggestions we could keep this post as sticky

Regards
Mirko
 
0ldman
Forum Guru
Forum Guru
Posts: 1446
Joined: Thu Jul 27, 2006 5:01 am

Re: As we would like to evolve the wireless

Tue Sep 17, 2013 7:03 pm

All good ideas.

I haven't tried it yet, but I was thinking about using WDS to sync as a work around.
Exactly, what you mean by wds sync?
Take two APs, set WDS on a virtual AP running under it that isn't attached to a bridge. Have the two VAPs connect to each other over WDS. WDS is keeping the two in sync, but not moving data back and forth. I don't know if this would work for a heavily loaded AP.
I was going to try this, but I haven't gotten around to it.
 
ejansson
Member
Member
Posts: 301
Joined: Fri Oct 21, 2005 4:09 pm
Location: Manitoba, Canada

Re: As we would like to evolve the wireless

Sat Sep 21, 2013 7:45 am

Filtering

Channel selectivity or reduction of adj sent channel interference. Most cards have poor filtering, improving this will make the cards perform bettering in environments with with high adjacent channel signal strength. The net effect is a reduction in S/N. Adding an external cavity filter will do the trick but you are then locked to that frequency. Providing good filtering will will give similar results with out the need for an external fixed freqency cavity filter.
 
ste
Forum Guru
Forum Guru
Posts: 1813
Joined: Sun Feb 13, 2005 11:21 pm

Re: AW: As we would like to evolve the wireless

Sat Sep 21, 2013 8:12 am

Filtering

Channel selectivity or reduction of adj sent channel interference. Most cards have poor filtering, improving this will make the cards perform bettering in environments with with high adjacent channel signal strength. The net effect is a reduction in S/N. Adding an external cavity filter will do the trick but you are then locked to that frequency. Providing good filtering will will give similar results with out the need for an external fixed freqency cavity filter.
There are minimal requirements defined in 802.x. Hope 802.11ac will raise this requirements?
 
telepro
Frequent Visitor
Frequent Visitor
Posts: 63
Joined: Sun Apr 03, 2011 7:50 pm

Re: As we would like to evolve the wireless

Sun Sep 22, 2013 10:30 pm

Would like to see option to configure a single wireless interface in CPE configuration to act simultaneously as both an access point for one SSID and as a station on another SSID. Useful where wired uplink is not available (thus station connection), and wireless often not Mikrotik components.
Thanks.
 
rborz
newbie
Posts: 29
Joined: Sat Nov 29, 2008 11:41 pm

Re: As we would like to evolve the wireless

Mon Sep 23, 2013 12:48 am

I've a script for this but changing the power/antenna gain cause a reconnect.
ste, this sounds interesting. Can you share your "ATPC" script?

Thanks,
Robert
 
Benik3
newbie
Posts: 46
Joined: Thu Apr 04, 2013 11:41 am

Re: As we would like to evolve the wireless

Tue Sep 24, 2013 12:55 am

Maybe also frequency usage scan to file :)
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Tue Sep 24, 2013 11:04 pm

My list is nearly the same ;-))
I would also like to see some sort of priorization in nv2. I would like to slow down cpes which have bad signal to save airtime for good installed cpes.
Good feature!! I add it to the list
I've a script for this but changing the power/antenna gain cause a reconnect.
If you want, post the script
MuMIMO in 802.11ac would be a great feature.
Explain the features of this technology

We should be able to see 90MBps in 10MHz Channel and 180MBPs in 20MHz with *optimal* conditions.
(pmp450 are very limited with SM in NATmode, guess it's a weak CPU).
You see 90Mbps in a single radio with 30-50 clients (PTMP) that traffic at the same time ? :shock:

We prefer solutions without external Coax-Cables. So integrated Antenna with box on the
back connected with short pigtails within the housing.
The new SXT Sector with a better housing/antenna would be a good solution.
SXT Sector has too low gain, but I think that the RB912UAG-5HPnD-OUT can be used on every type of antenna (45°,60°,90°,120°,360° OMNI, also on PTP link with a 30dB dish) Too many type of antenna to have an integrated product
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Tue Sep 24, 2013 11:11 pm

Filtering

Channel selectivity or reduction of adj sent channel interference. Most cards have poor filtering, improving this will make the cards perform bettering in environments with with high adjacent channel signal strength. The net effect is a reduction in S/N. Adding an external cavity filter will do the trick but you are then locked to that frequency. Providing good filtering will will give similar results with out the need for an external fixed freqency cavity filter.
Very good idea 8)
You mean something like the "Multi-Lane RF" technology made by ubnt?
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Tue Sep 24, 2013 11:14 pm

Would like to see option to configure a single wireless interface in CPE configuration to act simultaneously as both an access point for one SSID and as a station on another SSID. Useful where wired uplink is not available (thus station connection), and wireless often not Mikrotik components.
Thanks.
I think you can do this using WDS + VirtualAP...right?
WDS can operate like AP and a station
with VAP you can create virtual ssid
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Tue Sep 24, 2013 11:16 pm

Maybe also frequency usage scan to file :)
On my point one.. there are already these requests
 
User avatar
Aug
Member
Member
Posts: 313
Joined: Thu Jun 07, 2007 2:10 am

Re: As we would like to evolve the wireless

Wed Sep 25, 2013 1:46 am

Maybe "channel-width=Any"

I could use this on my 900Mhz cpe.
Aug
 
ste
Forum Guru
Forum Guru
Posts: 1813
Joined: Sun Feb 13, 2005 11:21 pm

Re: As we would like to evolve the wireless

Wed Sep 25, 2013 8:29 am

MuMIMO in 802.11ac would be a great feature.
Explain the features of this technology
It allows cpes to send at the same frequency at the same time. So it increases capacity.
We should be able to see 90MBps in 10MHz Channel and 180MBPs in 20MHz with *optimal* conditions.
(pmp450 are very limited with SM in NATmode, guess it's a weak CPU).
You see 90Mbps in a single radio with 30-50 clients (PTMP) that traffic at the same time ? :shock:
I have no 450. But heard this is possible when all cpes have good enough signal for highest modulation.
I guess this would be possible with optimized nv2, too. It is a polling protocol. Why should the amount of cpes reduce the capacity?
We prefer solutions without external Coax-Cables. So integrated Antenna with box on the
back connected with short pigtails within the housing.
The new SXT Sector with a better housing/antenna would be a good solution.
SXT Sector has too low gain, but I think that the RB912UAG-5HPnD-OUT can be used on every type of antenna (45°,60°,90°,120°,360° OMNI, also on PTP link with a 30dB dish) Too many type of antenna to have an integrated product
This external connectors add loss. Every single connector. So you might get better results with an integrated antenna using fewer different sector sizes.
I dont like external connectors as every connector has to be waterproof and is a point of failure.
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Wed Sep 25, 2013 9:23 am

It allows cpes to send at the same frequency at the same time. So it increases capacity.
it looks the same thing of GPS Sync :( I have to take a look at some explanation of this
technology
I have no 450. But heard this is possible when all cpes have good enough signal for highest modulation.
I guess this would be possible with optimized nv2, too. It is a polling protocol. Why should the amount of cpes reduce the capacity?
I agree with you.
Is only an optimization of the protocol that is missing.
In real conditions the cell (with many users) hardly exceed 25 Mbps of aggregate traffic
This external connectors add loss. Every single connector. So you might get better results with an integrated antenna using fewer different sector sizes.
I dont like external connectors as every connector has to be waterproof and is a point of failure.
So would you ask Mikrotik the creation of integrated antennas for each type of situation? How many and what models?
Or maybe we can ask a single model with angle of radiation (and gain) adjustable? similar mechanism of Airmax Titanium sectors?
 
0ldman
Forum Guru
Forum Guru
Posts: 1446
Joined: Thu Jul 27, 2006 5:01 am

Re: As we would like to evolve the wireless

Wed Sep 25, 2013 9:26 am

Maybe "channel-width=Any"

I could use this on my 900Mhz cpe.
/


/system script
add name=ChangeBand policy=ftp,reboot,read,write,policy,test,winbox,password,sniff,sensitive,api source=":local delais 40s\r\
\n:local liste 5mhz,10mhz,20mhz;\r\
\n \r\
\n:foreach band in=\$liste do={\r\
\n:if (![/interface wireless get wlan1 running]) do={\r\
\n/interface wireless set wlan1 channel-width=\$band wireless-protocol=any\r\
\n:delay \$delais\r\
\n}\r\
\n}\r\
\n"
\

/system scheduler
add comment="" disabled=no interval=40s name="schedcheckWifi" \
    on-event=ChangeBand start-time=startup
/system scheduler enable schedcheckWifi
/
The only problem is sometimes the script changes too often, gets stuck in a loop and won't connect. I set my watchdog address to the AP the radio connects to, if it gets stuck then it reboots and works fine. I'm sure the script can be tweaked to cure this, but I haven't done it yet. This is not mine, this is the work of myself and several other forum members put together.
 
ste
Forum Guru
Forum Guru
Posts: 1813
Joined: Sun Feb 13, 2005 11:21 pm

Re: As we would like to evolve the wireless

Wed Sep 25, 2013 12:31 pm

It allows cpes to send at the same frequency at the same time. So it increases capacity.
it looks the same thing of GPS Sync :( I have to take a look at some explanation of this
technology
No. It allows to decode 2 parallel signals at the same frequency at the same time.
GPS Sync tries to avoid parallel signals.
I have no 450. But heard this is possible when all cpes have good enough signal for highest modulation.
I guess this would be possible with optimized nv2, too. It is a polling protocol. Why should the amount of cpes reduce the capacity?
I agree with you.
Is only an optimization of the protocol that is missing.
In real conditions the cell (with many users) hardly exceed 25 Mbps of aggregate traffic
This external connectors add loss. Every single connector. So you might get better results with an integrated antenna using fewer different sector sizes.
I dont like external connectors as every connector has to be waterproof and is a point of failure.
So would you ask Mikrotik the creation of integrated antennas for each type of situation? How many and what models?
Or maybe we can ask a single model with angle of radiation (and gain) adjustable? similar mechanism of Airmax Titanium sectors?
It would be enough if they make a 60 and a 120 degree model.
Keep in mind. 60 does not end at 60 it has 3db less at 60.
I have very few places where I would place exactly 4x90s. I place Antennas where customers live and they never
distribute evenly around a tower.
 
Lakis
Forum Veteran
Forum Veteran
Posts: 701
Joined: Wed Sep 23, 2009 7:52 pm

Re: As we would like to evolve the wireless

Sat Sep 28, 2013 2:34 am

Mikrotik should add auto shutdown option like if u use battery when voltage rich 22v/11v should trigger shutdown mode or standby mode if its possible
- or can be implemented in UP models - on low battery should shutdown other equipment attached to device (exp. RB750UP)
- should add also Health graphs /tem.,volt.
- should add ability to recalibrate voltage sensor
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Sun Sep 29, 2013 1:55 pm

Mikrotik should add auto shutdown option like if u use battery when voltage rich 22v/11v should trigger shutdown mode or standby mode if its possible
- or can be implemented in UP models - on low battery should shutdown other equipment attached to device (exp. RB750UP)
- should add also Health graphs /tem.,volt.
- should add ability to recalibrate voltage sensor
Lakis, I opened this discussion to speak only of the improvements that relate to the wireless as you can read the first post (first line)
 
User avatar
janisk
MikroTik Support
MikroTik Support
Posts: 6283
Joined: Tue Feb 14, 2006 9:46 am
Location: Riga, Latvia

Re: As we would like to evolve the wireless

Tue Oct 01, 2013 11:39 am

things you can do already:
1) adjust scan list - enter frequencies only intended for the site, like 2 - 4 channels - super-fast reconnect even on 5GHz band
2) adjust access list rules on AP, there you can set minimal and maximal signal strength that is allowed to connect. On client RouterOS you can adjust connect list, if you have several APs to connect to.

3) lock down what data rates can client device use (RouterOS clients)
 
ejansson
Member
Member
Posts: 301
Joined: Fri Oct 21, 2005 4:09 pm
Location: Manitoba, Canada

Re: As we would like to evolve the wireless

Tue Oct 01, 2013 8:59 pm

1) 2-4 channels is not always practical in crowed areas, as you can't predict what channels will be free 6 months from now unless you are in an isolated area with out many other radios. If there is interference on all the channels you may not be able to reach some clients to give them a new freq list.

2) ACL are ok for small installation but not practical when you have hundreds or thousands of clients. Also in countries with winter you can have large fluctuation in signal strength not to mention the effect of wet snow or ice on trees.

3) Unless thing have change you can't lock client data rates if you are using NV2.
 
mennowz
Trainer
Trainer
Posts: 84
Joined: Tue Apr 09, 2013 8:50 pm
Location: The Netherlands
Contact:

Re: As we would like to evolve the wireless

Wed Oct 02, 2013 3:16 pm

Unless thing have change you can't lock client data rates if you are using NV2.
Works just fine :), locking data rates needs to be done @ the AP, i lock all clients to a MCS (and basic rate is just 6Mbps).

So in my case is use MCS11 .. clients get 120 or 6 Mbps.. somehow all clients prefer the 120Mbps option :lol:

Need screenshots? , just yell :D
Certified Mikrotik trainer in the BeNeLux!. http://www.mikrotiktraining.nl
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Sun Oct 06, 2013 11:40 pm

things you can do already:
1) adjust scan list - enter frequencies only intended for the site, like 2 - 4 channels - super-fast reconnect even on 5GHz band
2) adjust access list rules on AP, there you can set minimal and maximal signal strength that is allowed to connect. On client RouterOS you can adjust connect list, if you have several APs to connect to.

3) lock down what data rates can client device use (RouterOS clients)
Hi janisk is a pleasure to see your interest in the discussion!

In your opinion, could we have some of the changes in the first post in RouterOS?
Are technically possible?

By including these and other changes Mikrotik could become the best system, also in wireless world
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Sat Oct 12, 2013 12:57 pm

Advanced Scanning: Scans all visible proprietary APs and for each (respecting SSID, security settings, connect lists, etc.) connects and keeps track of the SNR, Rx/Tx signal strength, radioName, mac-address and distance.
In a normal scan the TX Signal Strenght and distance is not visible and the results show also all other operators APs

I add it on the list
 
shayanjameel08
just joined
Posts: 8
Joined: Sat Oct 12, 2013 10:02 am

Re: As we would like to evolve the wireless

Sat Oct 19, 2013 10:16 am

I haven't tried all the techniques yet, but I was thinking about using WDS to sync as a work around. nice web forum for nice information.
Last edited by shayanjameel08 on Mon Oct 21, 2013 3:40 pm, edited 1 time in total.
 
Lakis
Forum Veteran
Forum Veteran
Posts: 701
Joined: Wed Sep 23, 2009 7:52 pm

Re: As we would like to evolve the wireless

Sun Oct 20, 2013 11:02 pm

should be add
link load and nv2/nstreme capacity info in %
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Mon Oct 21, 2013 9:24 am

should be add
link load and nv2/nstreme capacity info in %
You mean, to see the real capacity of the wireless link and the percentage of its use?

For example:
Link Rx/Tx Max Capacity: 98Mbps/33Mbps
Link Rx/Tx Actual Load: 49Mbps/11Mbps
Link Rx/Tx Actual Load %: 50%/33%
 
Lakis
Forum Veteran
Forum Veteran
Posts: 701
Joined: Wed Sep 23, 2009 7:52 pm

Re: As we would like to evolve the wireless

Tue Oct 22, 2013 1:50 am

should be add
link load and nv2/nstreme capacity info in %
You mean, to see the real capacity of the wireless link and the percentage of its use?

For example:
Link Rx/Tx Max Capacity: 98Mbps/33Mbps
Link Rx/Tx Actual Load: 49Mbps/11Mbps
Link Rx/Tx Actual Load %: 50%/33%
Yes something like that or more like nv2/TDMA efficiency
I have something in mind how should be calculated if u have more clients on one AP
if some clients require more time slots for some amount of data than TDMA efficiency will be lower
less time slots per client for some amount of data better efficiency
It can be add nv2 efficiency TAB - list of actual nv2 efficiency separately for every client so it can help us to find most problematic clients
at the end of the list average nv2 efficiency

About Actual Load theoretical from data rate and CCQ, every p2p link know "his" throughput - it should be easy
so it can be add Actual Load %, Average 15min Load %, Average 1h Load, Average 1d Load etc etc
Sure u all asking Why this now? Well link throughput is not constant, in a crowded places in different time of day, traffic in air is different, interference is different so logical throughput is different so at the end we are looking for Average Load %.
Also
Should be add graphs for
1. "theoretical" or actual link throughput (24h)
2. nv2 efficiency (24h)
3. Average Load (24h)

Out of the topic it would be very lovely if MT can add temp. and voltage graphs but I guess its hard to be implemented or impossible

Thanks
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Wed Oct 30, 2013 9:25 pm

I forgot to add the functionality of the Wireless Controller
Wireless Controller feature allows to centralize wireless network management and data processing

Mikrotik has already drafted the technical specifications
http://wiki.mikrotik.com/wiki/Manual:Wi ... Controller

I add it to the list 8)
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Wed Nov 13, 2013 10:13 am

100-Employees.png
Can you put at least 2 of these 100 employees in the wireless industry?
Dedicated only to the development and improvement of wireless capabilities
You do not have the required permissions to view the files attached to this post.
 
User avatar
normis
MikroTik Support
MikroTik Support
Posts: 24272
Joined: Fri May 28, 2004 11:04 am
Location: Riga, Latvia

Re: As we would like to evolve the wireless

Wed Nov 13, 2013 10:23 am

100-Employees.png
Can you put at least 2 of these 100 employees in the wireless industry?
Dedicated only to the development and improvement of wireless capabilities
We would love to, but such people are really rare :)
No answer to your question? How to write posts
 
ste
Forum Guru
Forum Guru
Posts: 1813
Joined: Sun Feb 13, 2005 11:21 pm

Re: As we would like to evolve the wireless

Fri Nov 22, 2013 3:47 pm

We see the first vendor offering QAM256 for their PTMP-Equipment *now*.
It does GPS-Sync already. So MT please improve your wireless.

This promises up to 125MBit on a 20MHz Channel. Of course cpes need short clear LOS.
I have sites where this is possible ...
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Fri Nov 22, 2013 11:10 pm

Who is??
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Fri Nov 22, 2013 11:21 pm

100-Employees.png
Can you put at least 2 of these 100 employees in the wireless industry?
Dedicated only to the development and improvement of wireless capabilities
We would love to, but such people are really rare :)
take a couple of engineers Motorola

Who is, in mikrotik, the person that has developed the NV2 protocol?
He might be the perfect person
 
ste
Forum Guru
Forum Guru
Posts: 1813
Joined: Sun Feb 13, 2005 11:21 pm

Re: As we would like to evolve the wireless

Fri Nov 22, 2013 11:31 pm

Who is??
This is a MT Forum so I dont link to the competition.
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Fri Nov 22, 2013 11:39 pm

Who is??
This is a MT Forum so I dont link to the competition.
ok... I found it :)
 
ejansson
Member
Member
Posts: 301
Joined: Fri Oct 21, 2005 4:09 pm
Location: Manitoba, Canada

Re: As we would like to evolve the wireless

Mon Nov 25, 2013 7:36 pm

MT.

There are now several companies that are using 802.11 based chip with sync and what appears to have a better protocol then nv2 is right now. I have found a small company that has seems to have mastered the issues (although I have not tested) and does not begin with M or C... you should seriously consider Leveraging this technology via cooperation or licensing agreement. Be Glad to share with you as I would be VERY surprised if you have heard of them. Unfortunately the product are not available In Europe or North America at this time.
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Thu Feb 13, 2014 9:13 am

Dear Mikrotik, another 5 months have passed... :(
Have you start to develop some of the many changes required?

I don't know others, but wireless market of Wisp is ever expanding and increasingly, better tools are needed in order to compete against "enemies".

Please give me good news
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Tue Apr 29, 2014 12:14 pm

After a while some changes have been made:
Software
1 - Spectral Scan allow to save a file on the RB but they are working for a viewer to see this proprietary file http://forum.mikrotik.com/viewtopic.php ... 36#p423336
13 - Wireless Controller feature allows to centralize wireless network management and data processing. See http://goo.gl/wwYFqX ... Status: COMPLETE!
UPDATE 29/04/2014: Add the CAPsMAN feature http://wiki.mikrotik.com/wiki/Manual:CAPsMAN
14 - Auto frequency feature added by mikrotik that select the best channel to use ... Status: COMPLETE!

Hardware
2 - A dual-pol cpe (like Sexant) 23dB panel like this http://www.cyberteam.pl/en/produkt/show ... Dual_/MMCX , but with the ability to see the LEDs array for alignment use ... Status: COMPLETE!
UPDATE 29/04/2014: QRT-5 is ready for use http://routerboard.com/RB911G-5HPnD-QRT If possibile we need a low cost (Level 3 + NO_Gigabit version for customer's use)

I update the first post
 
User avatar
honzam
Forum Guru
Forum Guru
Posts: 2293
Joined: Wed Feb 27, 2008 10:27 pm
Location: Czech Republic

Re: As we would like to evolve the wireless

Tue Apr 29, 2014 12:55 pm

If possibile we need a low cost (Level 3 + NO_Gigabit version for customer's use)
+1
LAN, FTTx, Wireless. ISP operator
 
Lupin
Member Candidate
Member Candidate
Topic Author
Posts: 265
Joined: Mon Feb 16, 2009 10:22 pm
Location: Italy

Re: As we would like to evolve the wireless

Wed Aug 20, 2014 10:51 pm

Software
6 - 802.11ac Protocol is ready to use

Hardware
1 - NetMetal is coming out

Come on Mikrotik!!
You're great! Good Job

see other points
http://forum.mikrotik.com/viewtopic.php ... 40#p384640
 
odge
Member Candidate
Member Candidate
Posts: 102
Joined: Mon Nov 29, 2010 2:53 pm

Re: As we would like to evolve the wireless

Thu Oct 02, 2014 1:26 am

Hi

1) Surely 802.11r is must do for Mikrotik? In the tablet/education market, its a must, so you are losing all these sales to UBNT, Ruckus, HP and Cisco, because there is no fast roam support.
2) Support for mesh/WDS and CAPSman. You should simply set the AP once to join the WDS bridge. (you could also allow some sort of Virtual AP that gets opened with no security Access, and any unconfigured APs will by default try latch onto it, once on the network you can bring them onto the secure WDS network by sending the CAPSman settings to them as per normal. This would also make setup of WDS for dummies, and make the product very accessible to market (Seems like WDS is still beyond most people to actually setup, yet steps are simple and repetitive).


Any thoughts?
 
Niex
newbie
Posts: 37
Joined: Sat Jun 23, 2012 7:12 pm

Re: As we would like to evolve the wireless

Wed Dec 09, 2015 2:48 pm

Some news?
O would like to do a wireless scan like in nanostation, without disconnect, is that possible ?
 
ste
Forum Guru
Forum Guru
Posts: 1813
Joined: Sun Feb 13, 2005 11:21 pm

Re: As we would like to evolve the wireless

Wed Dec 09, 2015 4:02 pm

Some news?
O would like to do a wireless scan like in nanostation, without disconnect, is that possible ?
Not possible with current Atheros Chipset. You need a second Wireless chip like the nanostation ac (the 11n nanostation cant do it) or a chipset which can do this like Quantenna.

Who is online

Users browsing this forum: No registered users and 32 guests