Ed, It might be time to step back and talk about your goals. I thought we were talking about putting these stations on HamWAN, not creating a new independent network. I'm not opposed to helping you with this plan, but it certainly will be easier (and cheaper!) to lean on our existing infrastructure. I'd start by trying to get as many of the stations connected to Haystack as possible. I think it should work for all but 11. My experience with 2.4 GHz is that it's just really noisy. HamWAN settled on 5.9 GHz for a number of reasons: - ham-only, so relatively lower noise floor - equivalent size dishes will have significantly more gain on 5.9 GHz than 2.4 GHz. This gain happens to be about equal to the increased path loss, meaning you have net zero cost for moving up to the quieter band. - smaller Fresnel zone on 5.9 GHz allows mounting lower on towers and shooting through smaller gaps in trees (as long as there's a gap!) - down in the Part 15 section of the band, there's lots of spectrum for wide-channel, fast point-to-point links What criteria are guiding your triangle-topology design? What types of applications do you want to use this network for? What constraints have you been given (sounds like an in-house solution is preferred)? Tom On Sun, Mar 6, 2016 at 11:00 PM, Ed Morin <edmorin.jr@gmail.com> wrote:
Hi Tom,
Well, you have been busy! Thank you for looking into these things -- I know it can take a while sifting through the possibilities.
My "stick in the ground" model that I've been presently mulling is a "hub-and-spoke" sort of setup -- at least from a theoretical point of view because it sure doesn't resemble one on paper! (Owing to the geography of course...)
Core triangle of 12, 13, and 17 for reasons stated earlier -- UNLESS one of the other stations ends up having a "better" link Station 14 would hang off 12 (and possibly have a HamWAN link as well as it surprisingly has a possible shot at it) Station 16 would likely hang off 13 (and/or possibly) -- station 11 would have to hang off 16 (so having a redundant link to 16 would be worth considering) Station 18 is potentially a challenge, but might have a shot at 12 (and maybe HamWAN as well as you point out)
Anyway, we need to get on the roofs and "see what we can see" to get a better idea.
As an aid for this, I made a little "telescoping mast" out of some PVC pipe that I can put my phone on and hoist up 15 feet or so using a rope (from wherever I'm standing). (It's a Windows phone, so it's expendable. :-) I use it to first take a short 360 video. Once I have an idea of a potentially promising direction, I use a timer for taking a higher-resolution picture that I can study in more detail. The other day I was pleasantly surprised that I might very well have a good shot at station 16 from our home where there is a relatively convenient mounting point. (For testing convenience as it's a nice 3 km+ path.) Anyway, it was nice not having to drag out the extension ladder. :-) So, it may be helpful for scoping things out at the stations; we'll see. If any of you have ideas for simple site surveying, I'd love to hear them.
I don't know how this is all going to play out, but several folks on the ARES team are excited by the prospects and having some more hard data from a survey will definitely kick up the energy level I'm sure. :-) We're still kicking around ideas on what to do for inter-station linking; it can get expen$ive in a hurry... Has anybody here played with 2.4 GHz amps and dishes? They are relatively inexpensive and the choices for routers to use are plentiful and inexpensive as well... One of the ARES guys and I achieved (more or less) a 1.5 km link using Linksys routers with dd-wrt and only 70 mW into DIY helical antennas on flimsy tripods! It wasn't super-stable, but for only 70 mW, I thought it wasn't too bad...
On Sun, Mar 6, 2016 at 9:35 PM, Tom Hayward <tom@tomh.us> wrote:
On Sun, Mar 6, 2016 at 4:05 PM, Ed Morin <edmorin.jr@gmail.com> wrote:
As for locations. These are the fire station addresses and GPS coordinates that one of our other members put together:
Station Address City Zip Lat/Long 11 - HQ 8450 161ST AVE NE REDMOND 98052-3848 47.677913, -122.124938 12 4211 148TH AVE NE BELLEVUE 98007-3119 47.648426, -122.143646 13 8701 208TH AVE NE REDMOND 98053 47.680206, -122.063499 14 5021 264TH AVE NE REDMOND 98053-2718 47.651962, -121.987793 16 6502 185TH AVE NE REDMOND 98052-5039 47.664105, -122.093651 17 16917 NE 116TH ST REDMOND 98052-2246 47.703403, -122.114135 18 22710 NE ALDERCREST DR REDMOND 98053-5845 47.692245, -122.03717
Ed,
I have not done RF models for each of these, but some quick plotting on the map shows line of sight from Haystack to stations 12, 13, 14, 16, 17, and 18. Station 11 is in a low spot and the only opportunity I see is a direct link to station 16.
My thinking is to have a "core network" of links between stations 12, 13, and 17. Of all the stations, 13 seemed to be the most promising. Station 12 is practically next door to Microsoft's main campus and the noise level is huge there, but it potentially has great shots to several other stations which makes it attractive to having in the core. Station 17 has become somewhat of a "hub" station for ARES -- at least we continue moving in that direction; trees could be an issue there. One or two of the other stations might have coverage potential, but it's all showing even more spotty on the map than these others. (Of course if we were able to access a node on Cougar, everything changes for the better...)
Station 17 shows line of sight to station 12, 13 and 18, so could be somewhat useful as a hub, but not for all of the stations.
Keep in mind the coverage map on hamwan.org is binary and only shows signals greater than -70 dBm. This is essentially 100% signal strength. "Spotty" might mean 5 Mbps speeds instead of 15 Mbps. Of course it could also mean zero signal, especially if there are local issues not accounted for in the model, like trees. It's worth doing calculations for specific sites in those "spotty" areas with a tool like ubnt.com/airlink or Radio Mobile, and if it looks favorable, ask here on the mailing list for someone to come out and test. We've got a number of people here who love playing with our portable HamWAN rigs. :-)
Tom KD7LXL _______________________________________________ PSDR mailing list PSDR@hamwan.org http://mail.hamwan.net/mailman/listinfo/psdr
_______________________________________________ PSDR mailing list PSDR@hamwan.org http://mail.hamwan.net/mailman/listinfo/psdr