[PSDR] Spectrum allocation for point-to-multipoint sectorized cell sites published

[Bart Kus]

+PSDR List

The same page that has the generic ring also has the more realistic map, 
lower down.  It includes Scott's Paine Field site. Sugarloaf will likely 
be replaced by "Baldi" Mtn (Grass Mtn).  I'm not crazy about the Graham 
Hill ground bounce site.  Is there anything with elevation around there 
we can use to bridge Baldi+Capitol?  Given the spacing between the sites 
I don't think we need to worry about PtMP frequency mapping too much.  
The same sectors (azimuth+freq) can and should be used at each site.  
The only problem to solve would be the coordination of PtP link freqs 
between the sites, but that won't be a real problem until there's plenty 
of sites.  :)

The modems don't have automatic adaptive power, so if that's needed (and 
I don't think it will be), it would have to be coded by us.

There are no active sites right now.  I'm plowing through antenna 
testing as fast as I can before site deploys.  I'm working with Cougar 
Mtn mgmt to make that the first high site.  Then Paine Field will be the 
first long distance link of about 25 miles, very shortly after.  Do you 
have LoS to Cougar?

The question level is good.  It's exactly the stuff we need to be 
worrying about right now.  All the cyberspace/networking stuff can be 
changed from the comfort of our desks, but to change out antenna or 
radio hardware after it's been installed on a commercial tower is a much 
larger headache.

--Bart

On 2/3/2013 9:48 PM, Rob Salsgiver wrote:
>
> Ok. Just got done doing some catch-up reading on the site.....
>
> Do we have an updated "map" -- even on a theoretical basis?  I see the 
> generic ring on the site, and I know Scott Honaker had a few potential 
> center nodes filled in on a hand-sketch at one time -- I think mainly 
> because it was Snohomish DEM was looking to be one of the possible 
> inner link sites.
>
> The topic below begs for a map to better understand.  At first blush, 
> it would seem that the higher power amateur links would fit most 
> nicely with any longer distance backhaul requirements -- i.e. -- if we 
> had a link that needed to go from one end to the other North to 
> South.  Some of the lower power U-NII -- as you noted, for closer 
> links.  Once we start penciling in some starting frequency and channel 
> assignments -- even on a "theoretical" map, I think it will let us 
> "see" some of the more obvious problems and additional questions that 
> might need to be asked.  We may re-use U-NII frequencies on the north 
> and south ends, as long as they don't have any likely hood for direct 
> or even tropo RF shadows or ducting / overlaps.  These temperature 
> inversions we have can be a potentially big headache to deal with as 
> well.   Power might need to be adaptive based on rain or fog, or even 
> if we start to get indications of multipath.
>
> Let me know if we have a better map to work with, or if you want me to 
> start keeping notes and throwing some working materials together.
>
> Do we have any active high-sites at the moment for experimenting?
>
> Sorry if it's too much or too many stupid questions -- unfortunately 
> that's what you get when my brain starts following new rabbits down 
> the path...<g>
>
> Cheers,
>
> Rob
>
> *From:*PSDR [mailto:psdr-bounces at hamwan.org] *On Behalf Of *Bart Kus
> *Sent:* Sunday, February 03, 2013 2:38 PM
> *To:* psdr at hamwan.org
> *Subject:* [PSDR] Spectrum allocation for point-to-multipoint 
> sectorized cell sites published
>
> Hello,
>
> I've just published the proposed PtMP spectrum usage for HamWAN cell 
> sites.  It can be seen at the Spectrum Allocation 
> <https://www.hamwan.org/t/tiki-index.php?page=Spectrum+Allocation&structure=HamWAN> 
> page.  Optimized for viewing @ 2560x1600. :)  The point-to-point links 
> between cell sites will not be allocated/planned before hand, and 
> don't need to be published since no users will use those links.
>
> Any objections/suggestions?
>
> My biggest worry right now is when someone wants to use the ARC 34dBi 
> dish to link to a sector, but the dish's frequency response may suck 
> that high up.  The 21dBi grids seem to only lose 3dB of gain from 
> their peak at the very top of the spectrum.  Not awful.  Have not yet 
> measured the 27dBi grids, but expect similar results.
>
> One very important thing this plan does is identify the center working 
> frequency for the sector antennas as being 5.855GHz. That's the 
> frequency at which radiation pattern measurements should be made, and 
> is the very next task on my list.
>
> --Bart
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.hamwan.net/pipermail/psdr/attachments/20130204/a1ab32ac/attachment.html>