Hi,

I spent many hours on the code openwrt and I received the opportunity to use channels Between 2312 - 2407 MHz with numbering 237 to 255 like use AIrmax OS for superchannels.

Below result of command iw and iwinfo on my openwrt UBNT M2 to show channels available in my openwrt version. I have start from 2362 Mhz because in Europe it is first frequency on which we use digital communication for Hamnet but it is not problem to add start from 2312 Mhz with channel number of 237

========================================================================

root@HamNET:/# iwinfo
wlan0     ESSID: unknown
Access Point: 00:00:00:00:00:00
Mode: Client  Channel: unknown (unknown)
Tx-Power: 12 dBm  Link Quality: unknown/70
Signal: unknown  Noise: -86 dBm
Bit Rate: unknown
Encryption: unknown
Type: nl80211  HW Mode(s): 802.11bgn
Hardware: 168C:002A 0777:E012 [Ubiquiti NanoStation M2]
TX power offset: 12 dB
Frequency offset: none
Supports VAPs: yes  PHY name: phy0

root@HamNET:/# iwinfo wlan0  freq
2.412 GHz (Channel 1)
2.417 GHz (Channel 2)
2.422 GHz (Channel 3)
2.427 GHz (Channel 4)
2.432 GHz (Channel 5)
2.437 GHz (Channel 6)
2.442 GHz (Channel 7)
2.447 GHz (Channel 8)
2.452 GHz (Channel 9)
2.457 GHz (Channel 10)
2.462 GHz (Channel 11)
2.467 GHz (Channel 12)
2.472 GHz (Channel 13)
2.362 GHz (Channel 247)
2.367 GHz (Channel 248)
2.372 GHz (Channel 249)
2.377 GHz (Channel 250)
2.382 GHz (Channel 251)
2.387 GHz (Channel 252)
2.392 GHz (Channel 253)
2.397 GHz (Channel 254)
2.402 GHz (Channel 255)
root@HamNET:/# iw list|grep "MHz"
* 2412 MHz [1] (18.0 dBm)
* 2417 MHz [2] (18.0 dBm)
* 2422 MHz [3] (18.0 dBm)
* 2427 MHz [4] (18.0 dBm)
* 2432 MHz [5] (18.0 dBm)
* 2437 MHz [6] (18.0 dBm)
* 2442 MHz [7] (18.0 dBm)
* 2447 MHz [8] (18.0 dBm)
* 2452 MHz [9] (18.0 dBm)
* 2457 MHz [10] (18.0 dBm)
* 2462 MHz [11] (18.0 dBm)
* 2467 MHz [12] (18.0 dBm)
* 2472 MHz [13] (18.0 dBm)
* 2484 MHz [14] (disabled)
* 2362 MHz [247] (18.0 dBm)
* 2367 MHz [248] (18.0 dBm)
* 2372 MHz [249] (18.0 dBm)
* 2377 MHz [250] (18.0 dBm)
* 2382 MHz [251] (18.0 dBm)
* 2387 MHz [252] (18.0 dBm)
* 2392 MHz [253] (18.0 dBm)
* 2397 MHz [254] (18.0 dBm)
* 2402 MHz [255] (18.0 dBm)
root@HamNET:/# iwinfo wlan0 tx
* 12 dBm (  15 mW)
13 dBm (  19 mW)
14 dBm (  25 mW)
15 dBm (  31 mW)
16 dBm (  39 mW)
17 dBm (  50 mW)
18 dBm (  63 mW)
19 dBm (  79 mW)
20 dBm ( 100 mW)
21 dBm ( 125 mW)
22 dBm ( 158 mW)
23 dBm ( 199 mW)
24 dBm ( 251 mW)
25 dBm ( 316 mW)
26 dBm ( 398 mW)
27 dBm ( 501 mW)
28 dBm ( 630 mW)
29 dBm ( 794 mW)
30 dBm (1000 mW)
======================================

I have use Nanostation2 and Nanostation M2 to configure local hamnet network with Mesh node base on 802.11s

http://en.wikipedia.org/wiki/IEEE_802.11s 

https://github.com/cozybit/open80211s/wiki/HOWTO

instead OLSR

It is example how this work with NS2 and M2 UBNT:

root@OpenWrt:/#
Station 68:72:51:06:39:b1 (on mesh0)
inactive time:  20 ms
rx bytes:       15043
rx packets:     258
tx bytes:       8639
tx packets:     61
tx retries:     7
tx failed:      0
signal:         -24 dBm
signal avg:     -26 dBm
tx bitrate:     48.0 MBit/s
rx bitrate:     54.0 MBit/s
mesh llid:      933
mesh plid:      1578
mesh plink:     ESTAB
mesh local PS mode:     ACTIVE
mesh peer PS mode:      ACTIVE
mesh non-peer PS mode:  ACTIVE
authorized:     yes
authenticated:  yes
preamble:       long
WMM/WME:        yes
MFP:            no
TDLS peer:      no
root@OpenWrt:/# meshp
DEST ADDR         NEXT HOP          IFACE       SN      METRIC  QLEN    EXPTIME         DTIM    DRET  FLAGS
68:72:51:06:39:b1 68:72:51:06:39:b1 mesh0       8       171     0       4970    100     0       0x15
root@OpenWrt:/#

Merry Christmas

73 Waldek sp2ong

The big questions are:

How do these look on a spectrum analyzer?

Did you manage to extract cal data or are you just running these open loop? if you did what frequencies did you add for the control points and would you be willing to provide the code for the cal eprom reading files.

Are your running low power because the units have been seen to push the PA high as you go down in frequency or are you running it low just for testing?

Have you you tried these to see if there is any difference in RF modulation/demodulation performance as we push the chips out of band?

Nice work Waldek... you have our attention.  But there is another challenge we must overcome.  The Ubiquiti hardware is "tuned" by calibration data that is specific to each device.  This optimizes a device for the frequencies intended and brings the device into the manufacturer's published specifications.

We have seen write-ups similar to your approach, but no one has been able, yet, to show power output that hasn't been compromised by the manufacturer's calibration data (contained in CAL files).  We would be eager to hear that you have unlocked the mysteries of these CAL files and to see you confirm the power output of the Nanostation M2 on the 247-255 channels.

If this can be overcome, then your technique offers us enormous promise... particularly in the 3.4 GHz and 5.8 GHz bands.

Thanks,

Andre, K6AH

Hi Andre,

Yes I know, but I do not have the appropriate equipment to measure it.
Today I managed to change the ath5k driver and thanks to this old UBNT NS2 working well with 802.11s mesh. Attach a screenshot where a small network operates 802.11s, two devices  m2 and one ns2. Perhaps it 802.11s will be a better choice for me instead of OLSR, for this reason we will be able to use the NS2

I have try modify iw tools to display correct bandwidth channel, in current version iw-3.15 wehre I have use bandwidth 5 Mhz is displayed not correct

73 Waldek SP2ONG

Attachments
 mesh-2362Mhz.jpg [175 KB]

At current I have open on 2 Ghz frequency from 2312 Mhz to 2480 Mhz

Please look attached screen dump from UBNT NS2.

73 Waldek SP2ONG

Attachments
 channels.jpg [141 KB]

Hi Andre,

Can more tel me about CAL files ??? where I can find this.

I will try to check power output on channels below 2400 Mhz. When devices are located not big far all working OK.

73 Waldek SP2ONG

Hi Andre,

Thanks for info I will looking on this ART (Atheros Radio Test) (http://wiki.openwrt.org/doc/techref/flash.layout)

At current I can do very simple test show how is signal strength for a few different channels with display of SNR. Please look attached screen dump

Attachments
 pwr-channels.jpg [164 KB]

Hi Andre,

I thing that UBNT devices M2  will working good blow 2400 Mhz because in original firmware AirmaxOS in Compliance Test mode you can select channels 255-237 (see attached scren dump from AirMAXOS) and you can  see in my last post with attached screen dump where I show SNR and signal strength for different  channels

73 Waldek SP2ONG

Attachments
 airmax-ns2.jpg [103 KB]

You may want to try reaching out to Bart Kus, AE7SJ.  He is the guy behind hamwan.org and may have the necessary test equipment to look at and verify a modified openwrt build like we are talking about.

If anybody would like test firmware for UBNT M2 with open 2 Ghz bandplan the firmware is available :

http://hamnet.ugu.pl/download/openwrt-ar71xx-generic-ubnt-nano-m-squashfs-factory.bin

MD5: 77a39c6c9feb6427b8f6fb8cbc80b682

When you put the  firmware via TFTP you can obtain IP address via DHCP from range:

172.16.1.x

Connect to UBNT M2:

telnet to 172.16.1.1

set password use command:

passwd

next set current date and time UTC:

date -s '2015-02-01 15:00'

setup hamnet network use script

hamnet NODE-Name IP-address Net-Mask

for example:

hamnet KA2ABC-2 44.128.1.1 255.255.255.0

After reboot system  you can go to simple status page:

http://172.16.1.1:8080

and on this page click on penguin icon  to see OLSR status

default setup channel 247 2.362 Ghz and bandwidth 5 Mhz with olsr SSID "HamNET"

but you can change via uci or editor in /etc/config/wireless

This test version of FW don't have any WebGUI to setup.

73 Waldek SP2ONG

Attachments
 mesh-owrt-ns2.jpg [87 KB] :: Nanostation2 with AA v12.07
 mesh-owrt-m2-ms.jpg [42 KB]
 mesh-owrt-m2-stat.jpg [84 KB]
 mesh-owrt-m2-mn.jpg [84 KB]

This is probably the best explanation I have read on the calibration topic:

http://villagetelco.org/2009/11/rf-hacking/

Each AR2317 chip is a little different and consequently each chips requires individual RF calibration on the production line. This calibration ensures each AR2317 product has uniform transmit power and meets certain performance criteria. The calibration data is stored at the end of the SPI flash chip and loaded by the WiFi driver at boot time. Calibration is performed on the production line using a bunch of expensive RF test equipment connected via GPIB to a host computer running special software. The calibration procedure and most of the data relating to the AR2317 RF section is part of the Atheros “secret sauce”.....

The calibration data is stored on a part of the 8mb flash eeprom (around 64KB in size) which we never overwrite.

The Ham Bands do not always overlap the worldwide channel specs (For example HAM bands go well below on 2.4GHz and well above on 5.8GHz and in the case of 3GHz with the down converter the majority of the channels lie between the normal channels)   Are there specific channels that are Ham Band that you are thinking of that are in the global channel field?  I would guess you are thinking of a few 5.8GHz channels that are below the current channels that are actually in the USA band plan (but not supported officially by the USA model hardware by Ubiquiti due to legal certification needs) that partially overlaps the highly used US Government Radar band (FCC NAL letters go out often for against anyone in that frequency range with wifi that do not have DFS [we don't and we are not required to but we are required to not interfere] )  The main thing to stop those channels would be the lack of International model hardware board.info files as we would expect those devices to have the CAL data in them for sure. (Think we only have a Rocket board.info in XM series)

Today I have patched linux kernel 3.19 to open tx power range and frequency below 2.4 GHz on my UBUNTU laptop. I have use USB TL-WN722N , run olsrd and working with my local network on channel #247 2.362 Ghz. It is nice solution for mobile access to HamNET network.

I have tested other solution with use VirtualBox  where I have run my own OpenWRT and connected WN722N to laptop and working good with local HamNET network on 2.362 Gh. You can find how to use VirtualBox with openwrt and USB WIFI:

http://code.commotionwireless.net/projects/commotion/wiki/Virtual-Box-Real-Mesh

Attachments
 hamnet-wn722.jpg [90 KB]

Of course running with a wifi modem in laptop to a local mesh node is not recommended as it will congest the entire mesh network (your local mode "jams" the frequency as it repeats back to your laptop)

This block out can degrade an entire network for miles to hundreds of miles (depending upon the layout of the network)

It it is still recommended you either use an Ethernet cable or a wifi AP instead on a different channel.

I showed only possibilities using different equipment in Mesh Network but other issue is how to use in local network.

73 Waldek SP2ONG

I'm looking into the usefulness of re-crystalled WRT54s, to operate below 2400MHz
According to this German article (translated) it looks straight-forward. It appears similarly-mod'ed
units will interoperate, but raises questions as to its working with NS2 or TL-WN722, or other unit
in expanded frequency mode through software.

http://www.rlx.lu/~lx1tb/wrt54gs/

I suspect that scaling 20MHz clock to the WRT54 Wifi radio chip (BCM2050) will also
change the OFDM inter-carrier spacing  (nominally 312.5kHz) and other parameters;
unless these can be compensated by firmware changes. From what I read, OFDM
frequency error needs to be substantially less than the spacing, and TX-RX inter-carrier
spacing the same. Does anyone have additional knowledge/experience with this?

For anything which can be properly frequency scaled, the channel centre frequency
should fall very close to 23x2 or 23x7MHz eg. 2362.0; I believe within 100kHz or better.
That limits the choice of crystals. I was thinking of 19.440MHz instead of 19.6608 (which
is a baud-rate crystal and not likely to be as accurate/stable), or another common OTS
crystal 19.200MHz, but none of the 13 new frequencies "lands" close enough to target.
Frequency synthesizer was tried by author above but jitter rendered the link unusable.
I suppose a low-jitter PLL design is possible, if designed carefully.

So far I have not found a detailed spec for the BCM2050 online, to understand better
how it works. Can someone point to suitable documentation?

Dave   VE3KMV