You are here

Post by WL7COO - AirLink has proven to be very valuable in combination

27 posts / 0 new
Last post
KG6H
Post by WL7COO - AirLink has proven to be very valuable in combination
Dan, WL7COO, posted this elsewhere, but it is specifically relevant to the San Joaquin Valley sub-forum:

AirLink has proven to be very valuable in combination with Google Earth Pro.  

It looks like MAARO will be able to learn, test, train and demonstrate AREDN with an investment initially comprised of a single NanoStation on Big Bullion at 33' AGL, a single NanoStation on a Mobile Portable 40' push up mast and several smaller NanoStation go-box kits with parking lot or roof top deployable 10' masts.    

Development goals are focusing on the ability to provide demonstrations of broadband apps between any 2, 3 or 4 County Offices simultaneously.   I suspect that laptop based video conferencing,  even if just between two offices  at a time will be the clincher.  (... is MicroSIP  'heavy' multipoint capable?) 

Investing in only QTH and/or deployable kits that can be temporarily 'deployed' to any  Served Agency site in the County we should be able to effect adequate demonstrations of MESH capability and capacity to encourage eventual access to County or State owned Towers/ radio masts.

I am very fortunate to have a clear air shot at W6BXN on Little Bullion and also to our local W6MPA site on Big Bullion.  Both sites are well within a 22 degree beam width of my tower with clear Fresnel from about 40' up to the top plate at 95'.   Pending establishment of a 'backbone' link of just under 2 miles between these two Mtn Tops, this routing will connect Mariposa and Merced Counties.  

This should be adequate initial infrastructure for testing, training and demonstrating at as many sites simultaneously as the # of go-box kits and participating Amateurs permit.  Every Served Agency Location in the county except for the Fairgrounds, Public Works and the Sewage Treatment Plant but including the Hospital and CalFIRE MMU is located well within a 22 degree beam width from the site on Big Bullion with a NanoStation center focused on one at the Sheriff's Office Downtown.  

The 40' mobile deployable will be used to ground truth suitable locations for similar 3 person deployable Military Surplus 40' Tactical masts or equivalent.

With cautious planning, iterative acquisition and testing and a continuous feedback loop from Served Agencies, I think it is inevitable we'll see an escalating level of interest and support.  Once we've accomplished this I expect  implementing access to the remaining 3 County locations will become very simple. 

How is it going in Modesto?   Are any specific plans gelling for you yet?

I'll try to figure out how we get a 'CA SJV' Regional Forum going here.  
73
...dan wl7coo

KG6H
Modesto plans and other rumors

I believe Stanislaus ARES at this point is taking a "wait and see" approach to see how the TARC/Merced ARES deployment goes.  I'm not an officer with StanARES, but an active member, hence the subject title "... and other rumors."

Personally, I have purchased 3 NSM5 devices.  I purchased one just to check it out, and after downgrading the AirOS, I was able to install the latest AREDN software.  It runs great, but I have no other nodes to talk to from Modesto.  So I purchased 2 more and am awaiting their arrival in a week or so (via the North Pole from Antarctica, or something).  I am looking for where I can purchase compatible Bullets or Rockets with 90 degree sectors for hilltops.

At a minimum I am going to use the NSM5 for the Manteca ARC/K6MAN contest events to link all of our stations for logging.  We previously have used a single Wifi AP in a central location and it worked ok, but we had drop-outs from time to time.  Relying on built-in laptop Wifi antennas isn't the best when you're trying to cover a large area, so having a wired connection to each laptop from each NSM5, or if there is more than one laptop having a switch connecting the NSM5 to laptops, will make it so we have zero drop-outs.  I was asked to be the President for MARC last year, so for 2 years at least (office term) I'll be dedicating my Field Day and CQP to being at these locations and running the network.

I also have access to many Modesto-area sites with various towers, two Oso locations, and Ripon, so I am going to see what I can do there as I have permission to install anywhere so long as I get the as-build docs created.  At a minimum I want to get the best Oso location online with a Bullet or Rocket covering Modesto.  I'll put an NSM5 down at the main office and have Internet connectivity for Winlink and other ham related activities.  I'll also have an NSM5 up my QTH to Oso.  I'll offer a NSM5 node to StanARES at the EOC if they're interested and haven't done anything with Mesh at that time.

If TARC is interested, I would like to try to get a direct cross-valley yagi connection from Oso to Bullion to connect the SJV mesh.  At a minimum I want to get my island Internet-linked to TARC.  If Ripon Fire and/or MARC are interested, I can also get a low-level link from MID's Modesto location to MID's Ripon Gen and then to the MARC water tower which covers Manteca, but that is many steps away unless there are folks that want to take the initiative with those groups.

I also have plans to have my mini-Mesh interfaced with a Winlink Post Office and/or RMS gateway and configured such that even if the Internet is down, it will forward via any of the available VHF packet RMS gateways.  StanARES just had Ken KF6IDK come to our last training day and do a great presentation on RMS Express.  I'm hoping that its use takes off with StanARES for ICS message handling forms.  Once set up, RMS Express, basically an email client, seems much easier to pass info vs. the other methods.  As you may know, Internet, high-speed Mesh, Packet, and HF all work natively with RMS Express.

73,
Jason de KG6H

WL7COO
WL7COO's picture
Jason, thank you far sharing your thinking.

It sounds like you will single handedly create the kernel of a Modesto island.  Expect it to be used.

Working on the 'bigger picture' links like an OSO <==> Bullion link may take longer but once it gets done all of our local island efforts will be leveraged.

Please keep posting your progress, thinking and 'lessons learned'.  
Mini After Action Reports tend to be immediately useful for those making the same effort.

73
...dan wl7coo

KG6H
Oso <==> Bullion and other ramblings
Dan,
Yup, hopefully it will be used. I've seen some pretty cool applications that the OC Mesh folks have done, such as portable roadside nodes with cameras out on bike races, taking motion-activated pictures of bicyclists and their bib numbers with timestamps.

The airLink site is pretty amazing. Using a Rocket M5 (27dBm) and AM-5G-20-90 (20dBi 90degree sector) to a NSM5 (23dBm 16dBi 60degree sector) in Modesto at 5MHz shows 16.9Mbps/19.01Mbps speeds, at 10MHz 33.80Mbps/38.03Mbps, and at 20MHz 67.60Mbs/76.05Mbs. That's pretty amazing, and frankly, I don't see a need for over 10Mbps on these links, even for video calls, and certainly 30+Mbps would exceed the need for most ham/EMCOMM applications working simultaneously.

Using a pair of Rocket M5's and RD-5G30's (30dBi) between Oso and Bullion, the airLink site predicts 5MHz having 21.13Mbps speeds, 10MHz 42.25Mbps, 20MHz 84.5Mbps. Pretty phenomenal, and I'm curious to see if we can even achieve that distance (72 miles), but there is clear line of sight between the two locations and their both on peaks on both ends of the SJV. The airLink site shows even a pair of NSM5s can make a connection between the two sites - which I question, but it sure would be cool if it worked.

It's going to be interesting to see how we set up the sectors - I believe I'll have my first node zeroed in on Modesto. But depending on how well coverage works at the widest points, I may put up a second node and try to cover more north-ish to include Manteca or even more of San Joaquin County, and more east-ish to include Turlock. Time will tell, and my priority is Modesto.

Best news is that thanks to someone else on the forum here I've found $50 used Rocket M5 XM units on eBay. HeyWhatsThat is cool as well to show sectors and distances coverages.

Image Attachments: 
K6AH
K6AH's picture
Consider 3.4 GHz

You may want to consider 3.4 GHz as an alternative.  I have found 5.7 GHz to have too much commercial competition on mountain-tops.  Also, since there currently are no commercial allocation in the US on 3.4, the band is ultra quiet.  I have a 50 mile 3.4 GHz link operational in  San Diego county with 23-28dB SNR.

Andre

KG6H
3ghz vs 5ghz

Andre,
Even in the Amateur-only upper allocation of 5ghz?

Jason de KG6H

K6AH
K6AH's picture
That's certainly the right approach...
... and depending on the other Part 15 devices on the hill, it may very well work.  Time will tell if the new automotive short range communications service will end up with these channels and what impact they'll have on our technology.
KE2N
KE2N's picture
3.4

The "power user" of the 3.4 GHz segment is "radio location" which means RADAR and, these days, more sophisticated navigation technologies.  In many areas of the country I suppose it is quiet.  But if your site has visibility to any airport/seaport, you really should take a survey at 3.4 before spending a lot of money on 3.4 equipment. Here (overlooking Washington's Dulles airport) the upper half of the band - where the ARRL band plan has digital segment - is absolutely slammed with RF. I can receive quite strong signal levels with a NS3 in my basement, never mind up on the tower.
 

KG6H
PAVE PAWS
Good point, I think the site survey is wise. We do have PAVE PAWS which has required us to limit our 70cm repeater on Mt Oso to half watt. Where there is one military radar technology in use, there may be another. Additionally, we have 3 international airports within RF range: OAK, SFO, and SMF.
K6AH
K6AH's picture
I wouldn't assume there will
I wouldn't assume there will be an issue here.  I have also have an international airport and a major military installation directly in the 3GHz coverage area of my node.  
 
WL7COO
WL7COO's picture
Thank you for both addressing the elephant in the room:

If funding and time were not an issue, I'd purchase sufficient radios (2 ea) on all 4 bands and use them, one each at each Mtn top site for testing and the other in the MESHmobile to confirm relative Noise Level, Interference , etc  at all likely 'served agency' locations and the County Fairgrounds for all 4 supported AREDN freq bands.    

As things stand at the moment myself and at least a two others MAARO members are prepared to invest the  $s  and more importantly the energy it will take to study, sharing knowledge and lessons learned with each other in an effort to make it easier for others to accomplish the same thing to support something we very much believe is worthwhile.

Our most important  initial design goal is to make this opportunity easily accessible to any other local Amateurs by showing them why it is such a useful form of Amateur Radio as applied to Public Service and Emergency Communications vs undertake teaching them the underlying LAN/WAN technology or inferring such is a pre-requisite. This applies even more so to representatives of Served Agencies.  If we think of a Business Model, Amateurs by necessity are our first 'customers'.

I am 100% convinced that demonstrating what AREDN will do is the most effective way to elicit interest and willingness to participate both as an involved Amateur and as the reaction from representatives of Served Agencies.  The February 2016 Merced County ARES meeting clearly made that point to everyone in attendance.   Thank you again Ken, Rich and Spencer for collaborating to effect that demonstration and even more so for making it clear there is room around the AREDNien Tribal Campfire for all who wish to smell the magic smoke of technology leveraged collaborative effort.

We hope to begin ordering sufficient Ubiquiti radios to effect a robust and flexible Merced <=> Mariposa 'backbone' connection between the recently authorized TARC 5 GHz Rocket with its 20 db, 90 degree sector antenna pointed towards Merced and Atwater, via an intermediate hop to a similar (possibly smaller Nanostation or a 120 degree Rocket /Sector antenna) to a unit at the W6MPA site that has a clear air path over all of Mariposa's Government Center neighborhood in Downtown Mariposa  as well as the Hospital, MMU and the  Mariposa Fire Station etc  on Hwy 49 north of town as well as  ... wait for it ....  our weekly meeting location at Happy Burger.   We are hoping to place our initial orders for the backbone radios and antennas as well as our mobile and QTH components RSN.  All of these paths are less than 7.5 miles.

That said I'm optimistic whichever freq(s)  'best practice' suggests will work best around *here* will work sufficiently well to accomplish the goals of gaining adequate familiarity with AREDN for those of us actively pursuing it so we will soon have the capability to demonstrate it's real world use to virtually all of the potential 'Served Agencies' in the vicinity between their respective offices.   Once we are sufficiently confident from having tested capability at three or more  of our QTHs we'll undertake demonstrations at any interested Served Agencies & nearby ARCs while supporting any other interested Amateurs choosing to build similar nodes individually.

Our current short term goal is a Merced<==>Mariposa backbone as a  2/3rd MAARO funded build,  (I'll fund and retain the other 1/3rd on my tower after a direct W6BXN<==>W6MPA  is established),   a pair of individually funded AREDN Node MESHMobiles, each with a push up mast for use as  'Mid Mile Nodes'  and at least 2 individually funded,  QTH useable 'Go-Box' units are targeted to be our initial builds.   The MESHmobiles and QTH go-boxes design & integration testing will include 'customer' accessible 802.11n capability to support  in-QTH and  Served Agency/Amateur Radio Club Demonstrations.

This is the basic architecture/Implementation plan that I am very pleased to see emerging as MAARO 's vision. 

73
 ...dan wl7coo  

K6AH
K6AH's picture
A well thought out plan, Dan.
A well thought out plan, Dan.  And lacking any major local noise issues, I'm sure you will have more than adequate coverage and signal strength to successfully demonstrate the network's capabilities.

Andre, K6AH
KE2N
KE2N's picture
reference

It is possible the signals I am seeing are not from the Airport itself.

Here is a slightly-old document showing locations where one might expect to find radar in the 3.1-3.7 MHz band. It was done in relation to Fixed Satellite Service up at the higher end of the band.  Some of the MD and VA locations listed are within troposcatter distance of here.

Some of these radars are airborne and quite mobile.  PEP power of the airborne ones can be as high as 1 MW. Shipboard systems can be 4 MW (land-based systems are only 120 kW).  The antenna can have 40 dBi main beam gain. Hmmm.  The pulses are really short - measured in femto-seconds.

https://www.ntia.doc.gov/files/ntia/publications/ntia99-361.pdf
 

AE6XE
AE6XE's picture
I was reading up on radar
I was reading up on radar sites, since there is a NEXRAD station KSOX within 2 miles of my major nodes in the area--this is the Doppler weather radar in the LA basin.  They wait about a 1000x longer to receive the echo responses from the radar pulse.  These are in the 2Ghz band with possible harmonics in the 5Ghz band.

"the radar is only transmitting for about 7 seconds out of each hour--or 0.19% of the time. As such, the average output power of a NEXRAD site is 0.0019 * 750,000 = 1458 watts."

In the referenced document above in the section on mitigation techniques:

"Advanced digital signal processing techniques, such as forward error correction coding and bit interleaving, can be very effective in reducing the susceptibility of an FSS earth station receiver to EMI from adjacent-band radars and from other interference sources. "

This paper  references the 3Ghz radar specs:  pulse repetition of 1.125 to 6 kHz @ 1 to 51.2 Fsec pulses.   So if I'm doing the math right, this means it is transmitting < 1 sec each hour to put in perspective.   Given 802.11n is using these FEC and other techniques, leads me to believe that these radars won't be a noticeable source of interference to a mesh node.  Although I wonder if the echo returns would be strong enough...   

Joe AE6XE

 
KE2N
KE2N's picture
radar

I have included the spectrum I actually see on 3 Gz.  This is a solid wall of RF covering the digital segment of the band plan.  I am reminded, by the prominent center frequency of 3510, that there is a specific type of radiolocation that uses this frequency (and adjacent ones apparently).
http://www.ntia.doc.gov/files/ntia/publications/compendium/3500.00-3650.00_01MAR14.pdf
You can see from the peak in the center (rather than a small null) that this is not an OFDM data signal.  This type of radar uses  much longer pulses (micosecond) and is a more likely explanation for what I am seeing.  However, figure 1 in the document shows that I am in an "exclusion zone" - not sure I understand that. Seems more like target zone.

 

Image Attachments: 
WL7COO
WL7COO's picture
Using a spectrum Analyzer has always been something someone

else did.   Fluke Monitors provided everything I needed to know about cable and fiber signaling.  Wireshark or netmon made available the majority of the bit level data on up that were useful resolving a wide variety of issues at the lower network layers and massively expensive route and packet header logging enabled much of the necessary forensic data we occasionally needed to provide. (this was before we were aware that was probably already happening elsewhere and vastly more useable since it's occurrence was clearly disclosed to every user and acknowledged on signed dated documents)

I see that until I fully appreciate what you are sharing and where and how useful this (or some less expensively available subset) level of RF data will be for pioneering new MESH islands, particularly where Disaster Response  and Public Service deployments will require expeditious and functionally accurate identification of useful 'mid-mile' site locations.

Can you please share what equipment you use to acquire this data?  This may be an approaching 'next most useful' area of study.
I'm curious if this data can be accumulated while mobile along a gps tracked path for post processing and analysis on current production portable laptops or tablets (or Surface Pros if there is a distinction) and ... what additional hardware?

Thanks
...dan wl7coo

K6AH
K6AH's picture
The Ubiquiti AirOS provides a
The Ubiquiti AirOS provides a very useful RF spectrum analyzer.  The next time you sit down to flash a new node with the AREDN software, take a few minutes to check out the native tools in the manufacturer's operating system.  In fact, you may even want to dedicate a node to your toolkit specifically for this purpose.

Andre
WL7COO
WL7COO's picture
Andre, 'the next time' will also be the first time<g>

As a retired Federal Program Manager in an enormously underfunded Agency I learned to plan meticulously for what little Annual Authorizations I could depend on.  This is a tradition I've found valuable in many arenas, including optimizing our household budget <g>. 

I have sufficient confidence in AREDN's utility that I am speculating many 'Served Agencies', once aware of it's capabilities, won't be able resist mention of it & by doing so, to some extent,  committing to support it with accommodations in County, State and FEMA Grants pertaining to Emergency Communications and Planning.

73 & thanks
...dan wl7coo

KE2N
KE2N's picture
OEM

Yes this is a standard feature of the Ubiquiti OEM software. If you had one nano station M3 and one M5, a 12 volt battery and a lap top you are all set.  Using the standard Ubiquiti firmware, I would do the airView function shown above and snapshot the screen.  For 5 GHz I would also do the site survey function (20 MHz BW) and copy-paste the results into notepad on your laptop. Do it a couple of times and combine the results. I divide the 5 GHz band into an upper and lower half. There are too many channels to do all at once and see anything.  Afterwards you can analyze the data you collected.  I found I can import the text into a spreadsheet and make an Excel graph of signals vs frequency, where the size of the dot indicates the number of times that signal has shown up in the scans (attached).  These are only the access points, They each have one or more stations - those do not show up.  

I got my NSM3's from ubntshop who sell on eBay.  You have to fill out a form to import them - I think you are allowed to import 3 without any special dispensation, if you are a ham.
 

Image Attachments: 
kj6dzb
kj6dzb's picture
Aside from rfi... Richard
Aside from rfi... Richard wb6tae has documented how to build a off the grid hub running on solar. Our mesh record is around 300day. The airview tool in airos is a valuable tool. 73 kj6dzb
WL7COO
WL7COO's picture
Gentlemen, thank you to all three of you.

Learning that keeping a few non AREDN flashed Ubiquity NSMs or NSM locos around to use for RF assessment at possible Mid sites makes the investment iin a few seem imminent.

Watching the barnyard and pastures from my desk has some intrinsic value. We need a chicken cam, a ewe cam and a kennel cam, at least till we learn how to make good use of the NSM locos and the cameras and can make more elaborate use of them elsewhere.

I'll search the Forums for Richard's Solar info.  Time to sort how much power off grid Ubiquiti radios each need. 

Are there 'preferred' Ubiquiti sources?
How does one get 'Training Credits' to participate in Ubiquiti's  training program?

Thanks again and 73
...dan wl7coo

KG6H
Learning that keeping a few

Learning that keeping a few non AREDN flashed Ubiquity NSMs or NSM locos around to use for RF assessment at possible Mid sites makes the investment iin a few seem imminent.

Using the "AREDN U-Boot Test" one can backup the node before flashing to AREDN software, and one can restore the original AirOS.  It's fairly easy to move back and forth between versions.
 
WL7COO
WL7COO's picture
Thanks Jason. Restating the now obvious ....

Download AREDN U-Boot app and appropriate binary image(s), open box, plug & play POE & ethernet cables, run U-boot, save Ubiquiti image, install AREDN image, repeat for each Ubiquity you can lay your hands on.

To return a given radio to full AirOS functionality reverse above to restore previously saved image using U-Boot (... or tftp or PuTTY from any terminal session with access to the image?).

Thanks again & 73
...dan wl7coo

KG6H
Two backups
Yup.  Only thing is with any AirOS greater than 5.5, after your initial U-boot backup step, you have to downgrade from the shipped 5.6 to 5.5, then U-boot will be happy with a good/good test.  At this point, I would run another U-boot backup at this AirOS 5.5 state and use this as the image you would restore if you want to go from AREDN back to AirOS in the future.  You can then install the AREDN image via the AirOS 5.5 upgrade interface.
WL7COO
WL7COO's picture
A month later and ...... still learning more than bargained for.

After finding the Radio Mobile Chapter in 'Wireless Networking for the Developing World'  I have to say RM does seem to be a more universal and comprehensive link prediction package than airLink.  

RM Online offered more opportunities to succeed at researching specific concepts,  airLink seemed to be making it too easy and had a few very confusing settings.  (Market driven vs  Academic orientation?)  One cause of some confusion for me was;    'Link Margin'  in Ubiquiti's lingo is 'Fade Margin' in tech journals.  That airOS has an 'advanced setting' for 'Link Margin' seemed nonsensical till I understood it meant "What will you settle for as a Fade Margin".

Anyhow, AREDN.org is an unusually useful self education tool and I'm very thankful the community is thriving.

Thank you all for asking your questions, sharing your knowledge and expressing your opinions, in context,  that is all extremely useful.  

This community seems to be willing to work together to assist everyone getting past conceptual and comprehension stumbling blocks as expeditiously as  possible.  There is an unusually high degree of sensitivity and respect for everyone's 'tech' abilities and recognition that we're all contributing what we can to enrich and support all who wish to participate.

I am becoming increasingly convinced that to the extent we are able to,  we should provide fellow Amateurs with as easy a path into AREDN capability as possible.  Those who want to and can, will dig around under the surface to find out how it works, those who aren't interested or specifically equipped should have complete confidence there will always be help readily available.  I believe this mix will ensure a long and productive future.

I have a very  long way to go but I'm making sure the rope is left behind for others and the trail is very clearly marked.
Looking forward is exciting, looking backwards is satisfying.

'The smoke of this campfire is very good'

pax

 

K5DLQ
K5DLQ's picture
... one of my favorite posts
... one of my favorite posts of all time!

Thanks Pax!
K6AH
K6AH's picture
+1
+1

Theme by Danetsoft and Danang Probo Sayekti inspired by Maksimer