Securifi Community Forum
Securifi Products => Almond+ => Topic started by: ILikeFish on December 15, 2015, 04:48:40 pm
-
I'm trying to figure out what's going on with txpower on my Almond+, running the latest firmware.
In the UK, AFAICT 5GHz A-lower bands should be able to transmit at 200mW / 23dBm.
However, txpower is reported by iwconfig as 17dBm. I don't seem to be able to boost this.
Also, is there any plan to support DFS bands?
Thanks
Jon
-
Make sure you have your country/region set correctly. 8)
Also double check your 20/40/80Mhz setting as that limits the channels that can actually be used during DCS.
-
Yes, it's set to the UK. I have 80MHz selected, though adjusting it appears to have no effect on the txpower reported by iwconfig.
Thanks
J
-
Change it to 20Mhz and set the channel to 36-64 but those will top out at 20mW.
Changing to channels 100-140 ups the power.
Really the best thing to do is run with 80Mhz and let the system do it's thing to hunt up bandwidth. If you have a lot of interference (not seeing the throughput you are hoping for), then drop it to 20/40Mhz.
-
A quick look at my iwconfig. (U.S. specs)
Tx Power at 29 dBm or 800 mW
Running 80MHz, not a lot of interference in the area so it can hunt up the preferred channels with room. My 3x3 channel shows up nicely at 1.3 Gb/s. Note that my 5gig frequency is 5.745MHz putting me in the 800/1000mW range and the Tx Power is right there where it needs to be. Maybe post up your iwconfig output. (remove encryption and SSID info)
These guys did a nice basic writeup of what is possible in the US and Euro specs.
http://www.quantenna.com/pdf/Intro80211ac.pdf
sensitive info removed
wlan00 IEEE 802.11ac ESSID:"XXXXXX"
Mode:Master Frequency:5.745 GHz Access Point: XX:XX:XX:XX:XX:XX
Bit Rate=1.3 Gb/s Tx-Power=29 dBm
RTS thr:off Fragment thr:off
Encryption key:RRRR-DDDD-BBBB-NNNN-YYYY-XXXX-XXXX-XXXX [2] Security mode:open
Power Management:off
Link Quality=94/94 Signal level=-96 dBm Noise level=-95 dBm
Rx invalid nwid:795 Rx invalid crypt:0 Rx invalid frag:0
Tx excessive retries:0 Invalid misc:0 Missed beacon:0
wlan10 IEEE 802.11ng ESSID:"YYYYY"
Mode:Master Frequency:2.452 GHz Access Point: XX:XX:XX:XX:XX:XX
Bit Rate:450 Mb/s Tx-Power=29 dBm
RTS thr:off Fragment thr:off
Encryption key:RRRR-DDDD-BBBB-NNNN-YYYY-XXXX-XXXX-XXXX[2] Security mode:open
Power Management:off
Link Quality=94/94 Signal level=-96 dBm Noise level=-95 dBm
Rx invalid nwid:2301 Rx invalid crypt:0 Rx invalid frag:0
Tx excessive retries:0 Invalid misc:0 Missed beacon:0
-
I should also mention, given the title of this thread, that the power isn't "configurable" and shouldn't be. It is set according to the Standards and Specs of the given Region and Frequency/Bandwidth being used.
That said, without seeing the exact results of your iwconfig, can't help you much more but the US Region/Frequency Specs appear to be implemented correctly.
-
Thanks for the detailed response :)
My iwconfig output:
wlan00 IEEE 802.11ac ESSID:""
Mode:Master Frequency:5.18 GHz Access Point:
Bit Rate=1.3 Gb/s Tx-Power=19 dBm
RTS thr:off Fragment thr:off
Encryption key: [3] Security mode:open
Power Management:off
Link Quality=94/94 Signal level=-96 dBm Noise level=-95 dBm
Rx invalid nwid:2354 Rx invalid crypt:0 Rx invalid frag:0
Tx excessive retries:0 Invalid misc:0 Missed beacon:0
wlan10 IEEE 802.11ng ESSID:""
Mode:Master Frequency:2.462 GHz Access Point:
Bit Rate:216.7 Mb/s Tx-Power=20 dBm
RTS thr:off Fragment thr:off
Encryption key: [3] Security mode:open
Power Management:off
Link Quality=94/94 Signal level=-96 dBm Noise level=-95 dBm
Rx invalid nwid:2752 Rx invalid crypt:8201 Rx invalid frag:0
Tx excessive retries:0 Invalid misc:0 Missed beacon:0
It's currently on channel 36 (there are only 4 channels available, and IIRC 80MHz will cover them all), country is UK and there are zero other 5GHz networks where I am... Max permissable power is 200mw, which is around 23dBm IIRC.
I'll try changing it to 20MHz and see if that's any different.
J
-
I get the same in the UK - txpower on 5GHz maxes out at 19dBm (~80mw) of the allowed 23dBm (~200mw). No matter what channel width I try.
I'm also wondering if DFS support will ever arrive - allowing us to use the 30dBm (1000mw) channels (100-140).
My 7 year old draft-n spec Airport Extreme handles DFS channels just fine..
-
Thanks for the detailed response :)
My iwconfig output:
wlan00 IEEE 802.11ac ESSID:""
Mode:Master Frequency:5.18 GHz Access Point:
Bit Rate=1.3 Gb/s Tx-Power=19 dBm
RTS thr:off Fragment thr:off
Encryption key: [3] Security mode:open
Power Management:off
Link Quality=94/94 Signal level=-96 dBm Noise level=-95 dBm
Rx invalid nwid:2354 Rx invalid crypt:0 Rx invalid frag:0
Tx excessive retries:0 Invalid misc:0 Missed beacon:0
wlan10 IEEE 802.11ng ESSID:""
Mode:Master Frequency:2.462 GHz Access Point:
Bit Rate:216.7 Mb/s Tx-Power=20 dBm
RTS thr:off Fragment thr:off
Encryption key: [3] Security mode:open
Power Management:off
Link Quality=94/94 Signal level=-96 dBm Noise level=-95 dBm
Rx invalid nwid:2752 Rx invalid crypt:8201 Rx invalid frag:0
Tx excessive retries:0 Invalid misc:0 Missed beacon:0
It's currently on channel 36 (there are only 4 channels available, and IIRC 80MHz will cover them all), country is UK and there are zero other 5GHz networks where I am... Max permissable power is 200mw, which is around 23dBm IIRC.
I'll try changing it to 20MHz and see if that's any different.
J
You need to be on a different channel. The reported frequency only supports 20mW.
UK/Japan/Eur spec iirc
Ch 36-64 20mW
Ch 100-140 1000mW
80 Mhz grabs up 4 20MHz Channels at a time, 5 total available in the U.S., only 4 in the UK / Japan.
You want to be at 5490-5710MHz and you are currently at 5.180MHz
There is a new spec for Europe for 2015 but the Almond+ hardware/radios are a couple years old and I would review the older spec.
New Spec
http://www.cisco.com/c/dam/en/us/solutions/collateral/enterprise-networks/802-11ac-solution/q-and-a-c67-734152.doc/_jcr_content/renditions/q-and-a-c67-734152_9.jpg
-
I get the same in the UK - txpower on 5GHz maxes out at 19dBm (~80mw) of the allowed 23dBm (~200mw). No matter what channel width I try.
There are a number of possible reasons for this including beam forming.
I'm also wondering if DFS support will ever arrive - allowing us to use the 30dBm (1000mw) channels (100-140).
My 7 year old draft-n spec Airport Extreme handles DFS channels just fine..
Pros and Cons of DFS are numerous. Lots of non commercial systems do not implement this due to the requirement to scan for radar and shutdown for 30 mins if found for example.
-
I'm also wondering if DFS support will ever arrive - allowing us to use the 30dBm (1000mw) channels (100-140).
My 7 year old draft-n spec Airport Extreme handles DFS channels just fine..
Just a little more info on DFS and why it's more of challenge to implement.
snipped from http://www.extremenetworks.com/understanding-fccetsi-dfs-regulations-for-802-11ac-deployments
Before getting carried away with the higher data rates available with 802.11ac it is important to understand the regulatory aspects for the 5GHz band. Many of the frequencies in the 5GHz band are also used by radar equipment for first responders, airports, weather stations and military installations. Because of this there are very stringent government regulatory requirements that must be followed by Wi-Fi radios when operating on these frequencies. The regulatory bodies specifying and enforcing these requirements in North America is the Federal Communications Commission (FCC) and in the European Union it is the European Telecommunications Standards Institutes (ETSI). Requirements/specifications for operating on the 5GHz band are called Dynamic Frequency Selection (DFS). DFS allows unlicensed devices to use the 5GHz band already allocated to radar systems without causing interference to those systems.
These DFS requirements are complex and costly to get regulatory certification from FCC/ETSI. Following are the typical requirements for access points (AP) operating on 5GHz DFS channel:
Before starting operation on DFS channel, scan the channel for Radar devices for 1 minute. If radar is detected follow step 4 otherwise start operation of DFS channel.
Must detect non-Wi-Fi Radar devices with pulse width as small as 5us
When radar device are detected stop operation on the channel within 500 milliseconds.
For AP vendors, this also means informing wireless clients to move away from this channel.
Do not become operational of this DFS channel for at least 30 minutes and after that go to step 1 before becoming operational again.
Given the above complexity many small office-home office vendors as well as some enterprise AP vendors don’t allow operation on DFS channels in the 5GHz band at all. Many client devices are also not certified for DFS band operation. Even when APs are certified and support DFS channels AP vendors often recommend operation exclusively on non-DFS channels given the uncertainty of DFS channels support at the client device level.
-
You need to be on a different channel. The reported frequency only supports 20mW.
200mW actually, but yes. However, 19dBm is just under 80mW. This is what I don't quite understand, why is it apparently limited to a low tx power? You mention beamforming in a post above?
UK/Japan/Eur spec iirc
Ch 36-64 20mW
Ch 100-140 1000mW
80 Mhz grabs up 4 20MHz Channels at a time, 5 total available in the U.S., only 4 in the UK / Japan.
You want to be at 5490-5710MHz and you are currently at 5.180MHz
That's kind of the problem - we can't, as those bands in the UK require DFS. That's why I'm interested in maximizing tx power on the A-lower bands, as those are the only ones that don't need DFS.
Thanks!
Jon
-
200mW actually, but yes. However, 19dBm is just under 80mW. This is what I don't quite understand, why is it apparently limited to a low tx power? You mention beamforming in a post above?
That's kind of the problem - we can't, as those bands in the UK require DFS. That's why I'm interested in maximizing tx power on the A-lower bands, as those are the only ones that don't need DFS.
Thanks!
Jon
I'll have to do a little digging tomorrow, out of time today. But I do remember that there were changes proposed to the Euro spec because Tx Power was not as good as it could have been because of specs related to aggregated 20MHz channel spectral masks versus considering the 80MHz bandwidth as a single channel and changing the allowances/tolerances.
http://www.ieee802.org/11/email/stds-802-11-reg/msg00354.html
My gut tells me I'm overthinking this a little and the answer is more straightforward, such as what devices are connecting to your network.
Silly question, do you have any 802.11ac capable devices that support 3x3, like a 2013 or later MACBook Pro?
-
I'll have to do a little digging tomorrow, out of time today. But I do remember that there were changes proposed to the Euro spec because Tx Power was not as good as it could have been because of specs related to aggregated 20MHz channel spectral masks versus considering the 80MHz bandwidth as a single channel and changing the allowances/tolerances.
http://www.ieee802.org/11/email/stds-802-11-reg/msg00354.html
I don't understand a word of that ;-)
My gut tells me I'm overthinking this a little and the answer is more straightforward, such as what devices are connecting to your network.
Silly question, do you have any 802.11ac capable devices that support 3x3, like a 2013 or later MACBook Pro?
I'm typing this on a Late 2013 Macbook Pro (MacBookPro11,3) which definitely supports 802.11ac, not sure about 3x3 though - I'll look it up.
Having said that, I can't get the macbook to stay working on 5GHz with my Almond+ for more than a minute or so - wifi stays connected but no IP packets seem to make it through. That's probably another question though!
J
-
There are several well known issues with MACBook Pro's and 802.11ac and 802.11h and some workarounds.
I had the links back before the site went down, will have to hunt them up again.
-
This is of course config on the Client side.
The goal being to optimize the client side and check throughput as you are configured to get 1.3Gb/s. The Tx Power setting should extend your range a bit given that everything else is optimal. Thing is, with the MACBook Pro, not everything is optimal.
https://discussions.apple.com/message/27858756#27858756
https://discussions.apple.com/thread/6825371?start=0&tstart=0
Going to a couple quick test on my A+, change the region and take a look at the results.
-
This is with auto channel select.
wlan00 IEEE 802.11ac ESSID:"xxxx"
Mode:Master Frequency:5.18 GHz Access Point: xxxxxx
Bit Rate=1.3 Gb/s Tx-Power=19 dBm
RTS thr:off Fragment thr:off
Encryption key:xxxxx [3] Security mode:open
Power Management:off
Link Quality=94/94 Signal level=-96 dBm Noise level=-95 dBm
Rx invalid nwid:2 Rx invalid crypt:0 Rx invalid frag:0
Tx excessive retries:0 Invalid misc:0 Missed beacon:0
wlan10 IEEE 802.11ng ESSID:"xxxx"
Mode:Master Frequency:2.452 GHz Access Point: xxxxx
Bit Rate:450 Mb/s Tx-Power=29 dBm
RTS thr:off Fragment thr:off
Encryption key:xxxxx [3] Security mode:open
Power Management:off
Link Quality=94/94 Signal level=-96 dBm Noise level=-95 dBm
Rx invalid nwid:52 Rx invalid crypt:0 Rx invalid frag:0
Tx excessive retries:0 Invalid misc:0 Missed beacon:0
-
And then changing the Region to Brazil and selecting Ch 161 / 80MHz
wlan00 IEEE 802.11ac ESSID:"xxxx"
Mode:Master Frequency:5.805 GHz Access Point:xxxxxx
Bit Rate=1.3 Gb/s Tx-Power=29 dBm
RTS thr:off Fragment thr:off
Encryption key:xxxxx [2] Security mode:open
Power Management:off
Link Quality=94/94 Signal level=-96 dBm Noise level=-95 dBm
Rx invalid nwid:22 Rx invalid crypt:0 Rx invalid frag:0
Tx excessive retries:0 Invalid misc:0 Missed beacon:0
wlan10 IEEE 802.11ng ESSID:"xxxx"
Mode:Master Frequency:2.452 GHz Access Point: xxxx
Bit Rate:216.7 Mb/s Tx-Power=29 dBm
RTS thr:off Fragment thr:off
Encryption key:xxxxx [2] Security mode:open
Power Management:off
Link Quality=94/94 Signal level=-96 dBm Noise level=-95 dBm
Rx invalid nwid:36 Rx invalid crypt:0 Rx invalid frag:0
Tx excessive retries:0 Invalid misc:0 Missed beacon:0
So the radio and drivers are doing their thing properly. Tx Power is up. Getting about 800mW.
This is of course not a solution, use the Region appropriate to your location. So back to taking a look at the limitations on the UK Region implementation.
-
Also recommend you load up an app called WiFi Explorer (free) and check out your network.
In the 5gig stuff, under Advanced Details you will see something called VHT Capabilities and can expand it to see VHT Capabilities Info.
-
Creator SECURIFI on February 5, 2015
@Rhys - In supports explicit transmit beamforming, but not implicit beamforming which we guess it what you're asking for. The Almond+ can not be upgraded to support 600Mbps on 2.4GHz as the chipset we're using doesn't support it.
-
Only other thing coming to mind is I believe, and am trying to confirm, that the 802.11ac spec for Europe was updated in June of 2015.
I know Wave 2 products are starting to hit and I'm wondering if the driver needs some tweaks for the UK/Eur settings, thinking it does.
Also, check System and Kernel logs. I am actually getting Radar detected (meaning some on that frequency) and the system is making adjustments.
None of this helps obviously, if the driver hasn't been tweaked to turn on the support for the other channels. Just can't figure out why it wouldn't be yet.
Last thing to try is to manually turn on some options related to supported channels and power but I don't believe we can even at the command line level. Might be able to compile a util that will do that if the hooks are there. If I'm really lucky the binary is already out there for our OS level/config.
-
I'd have to agree with your assessment that its the configuration of the regulatory domain that is likely the issue.
I've quite successfully had 29dBm out of it if I set the reg domain to USA.
Is the regulatory info built into the driver? Seems like a lot of effort to go through to just update something that might be updated multiple times a year given rules could change anywhere in the world.
I had thought it was /usr/lib/crda/regulatory.bin (dated Dec 2014 on my Almond+) but I tried replacing it with a custom-built one containing the correct radio levels and it made no difference.
-
I'd have to agree with your assessment that its the configuration of the regulatory domain that is likely the issue.
I've quite successfully had 29dBm out of it if I set the reg domain to USA.
Is the regulatory info built into the driver? Seems like a lot of effort to go through to just update something that might be updated multiple times a year given rules could change anywhere in the world.
I had thought it was /usr/lib/crda/regulatory.bin (dated Dec 2014 on my Almond+) but I tried replacing it with a custom-built one containing the correct radio levels and it made no difference.
We don't have iw or regdbdump
The bin file has to be properly signed with public & private keys and it depends on whether crda was built with openSSL or not, if not then we can't read the public keys.
But it shouldn't be a big deal to compile everything with the current db and push it out with a firmware release. Thinking the UK folks just need to make some noise. Not sure if we would be able to load everything needed to make the minor change and update it ourselves but it's not that tricky.
-
I'd have to agree with your assessment that its the configuration of the regulatory domain that is likely the issue.
I've quite successfully had 29dBm out of it if I set the reg domain to USA.
Is the regulatory info built into the driver? Seems like a lot of effort to go through to just update something that might be updated multiple times a year given rules could change anywhere in the world.
I had thought it was /usr/lib/crda/regulatory.bin (dated Dec 2014 on my Almond+) but I tried replacing it with a custom-built one containing the correct radio levels and it made no difference.
How did you build the custom file? Is everything there for a local compile/build?
-
How did you build the custom file? Is everything there for a local compile/build?
If it was and you did and that failed, then the stuff is hardcoded and we have to put a bogus Country code in for the UK and recompile.
-
If it was and you did and that failed, then the stuff is hardcoded and we have to put a bogus Country code in for the UK and recompile.
Which of course I do not suggest, condone or otherwise recommend. 8)
-
Instructions I used to make my own regulatory.bin are HERE (http://networkgeekstuff.com/networking/how-to-create-custom-linux-wi-fi-regulatory-database-to-unlock-30db1000mw/)
There are a few broken links but you can find the packages easily enough.
I used an ubuntu server box to create a signed regulatory.bin - I used the one present on my Almond+ as a starting point which I SCP'd off to ubuntu, then used regdbdump to dump into a text file which I edited the parameters for UK, then used the files in the crda package to compile it back into a regulatory.bin and resign it.
I then replaced the original regulatory.bin with my own followed by a full reboot. I'm assuming (but don't know enough really to say for sure) that if I'd created an invalid regulatory.bin and Almond+ was really using that file to set the reg info, that it would have broken wireless (or somehow otherwise impacted things) rather than continued working exactly the same way. If my changes were correct, and the regulatory.bin was correct, presumably it would then have listened to my changes.
Nothing happened at all, so my conclusion is that the reg info is being set elsewhere, possibly as you highlighted, within the compiled driver itself. This would be unfortunate as there is NO way to fix that ourselves - we are completely reliant on Securifi to have the inclination to build and test new drivers.
Given the rate of updates over the last few months I'd be surprised if we saw this update any time soon unfortunately :(
-
That's the how I do several things, build it over on another box (often ubuntu) and push it to the Almond.
As long you have the keys working properly, it should be fine.
btw, that's the same link (networkgeekstuff) I bumped into yesterday when I was looking through this stuff haha
The only thing to note is creating your 'custom country' code versus changing the info for the UK. If you went the custom country code route, I could understand why that would fail.
Meaning what might be hardcoded is the use of Eur/UK stuff and so you have to actually change the Eur/UK settings instead of telling the box to use a different Region. Even if they did that, it shouldn't be hard to change. Just have to dig up that other link that showed a sample of that instance.
I have to go back to the Kickstarter and see how they shipped things to different regions.
Still the best solution is to have Securifi update the Region info. Unless there's some other problem we're not privy to, I don't think it's a lot of work.
-
I did indeed edit the GB (its Great Britain rather than UK for this file) parameters.
No difference. Nothing broke, nothing changed - leads me to believe regulatory.bin is not in use on Almond+.
Hopefully someone from Securifi could comment here and give us a hint of some kind that this can be remedied!
-
Having been through a bit of back and forth with Securifi support about this - including a factory reset, and them testing on their own Almond+ units, they confirm the issue is to do with the driver for the wireless - regulatory info is baked into it.
They say that:
the transmitting power of 5 GHz completely depends upon the region and also the driver which we receive from the hardware manufacturer
and
unfortunately, we would not be able to do much on this. I could only request your understanding on this.
To be honest, that's less helpful than I was expecting (maybe I'm expecting too much?) - presumably they are saying this as the wireless driver we currently have is never going to be replaced to correct this issue, because they either can't or won't update it.
Not even a 'we're not working on it at the moment but its coming'.
Sigh.
-
Simple solution, change to US, use the UK frequencies, problem solved.
It's not illegal to set your router to the US region, as long as you stick with the frequencies allowed in the UK.
-
Already tried that - was the first thing I did in fact when I initially saw this thread.
Sadly the channels available on US reg zone are 36-48 and 149-161.
Under channels 36-48 the USA-reg txpower is limited to 17dBm (vs 19dBm on UK reg zone), and UK reg zone only covers up to channel 140, and anything above channel 48 requires DFS, which is not supported by the A+ at all.
I've tried a few reg zones trying to find a zone that allows better output but I'm yet to find one.
Basically, with no DFS support, I can only *ever* use 36-48 in the UK. The UK reg domain on A+ only allows 19dBm on those channels. USA only allows 17dBm, Russia only 17dBm, and I've tried a few others - none I've tried to use give more txpower on channels 36-48.
If you've got any ideas on zones that might allow better txpower on 36-48, I'll give them a try but I don't hold out much hope.... :'(
-
Having been through a bit of back and forth with Securifi support about this - including a factory reset, and them testing on their own Almond+ units, they confirm the issue is to do with the driver for the wireless - regulatory info is baked into it.
They say that:
and
To be honest, that's less helpful than I was expecting (maybe I'm expecting too much?) - presumably they are saying this as the wireless driver we currently have is never going to be replaced to correct this issue, because they either can't or won't update it.
Not even a 'we're not working on it at the moment but its coming'.
Sigh.
Securifi was mistreated or mislead by the original chipset vendor imho. This is one of the symptoms of that relationship. Nothing they can do about it now after the buyout. There has been such an investment into this SoC, Securifi is hamstrung unless Realtek steps up to help out and the economies of scale/ROI just aren't there for a dated chipset.
-
Securifi was mistreated or mislead by the original chipset vendor imho. This is one of the symptoms of that relationship. Nothing they can do about it now after the buyout. There has been such an investment into this SoC, Securifi is hamstrung unless Realtek steps up to help out and the economies of scale/ROI just aren't there for a dated chipset.
That doesn't have anything to do with the Wi-Fi, just everything else related to the Almond+...
DFS isn't supported by a lot of routers as 1. it's hard to do 2. it's not well supported by the chipset makers 3. it's hard to certify for.
As for using the DFS channels, it's highly unlikely anyone is going to come around knocking, so I wouldn't worry too much about it. You're right though, the US band doesn't support the EU frequencies, so you might have to try some alternatives.
This might help https://en.wikipedia.org/wiki/List_of_WLAN_channels#5.C2.A0GHz_.28802.11a.2Fh.2Fj.2Fn.2Fac.29.5B17.5D