[TowerTalk] "The Socialized Power Grid"

[Jim Lux]

> >
> > For hams, there's another alternative.  I know of several people
seriously
> > working on putting very high speed digital backbones on mountaintops
> > (where
> > there are existing repeater sites).  Originally it started as a "better
> > way"
> > to link repeaters, but then they took a step back and realized that
given
> > the relatively fixed infrastructure costs, they might as well go to much
> > higher rates, work with it as a fairly vanilla IP network, and use
> > techniques like VoIP for the linking.  Even a 200km 802.11g (2.5 GHz
> > 54Mbps)
>
> VoIP is a shaky proposition.  There are ISPs and even backbone operators
> that give it lower priority when loads are heavy.

Not VoIP via the Internet, but VoIP on the "private backbone" operated by
the hams.  They could assign whatever priority they care to.  Presumably,
since the network topology is "known", a lot of the latency and uncertainty
things that plague commerical internet VoIP wouldn't be an issue.

  There are also problems
> that extend beyond the interoperability issues and propritary protocols.
> I'd expect VoIP to become a premium service before long and the VoIP
> companies are going to start paying through the nnose due to the bandwidth
> as it catches on.
> However VoIP is a protocol that sends voice over existing internet setups,
> it is not a stand alone system.

Over IP networks, not only the Internet.  It's just a way to push voice
traffic using IP packets over whatever bearer you want to use.

>
> > link isn't out of the question with relatively simple off-the shelf
> > hardware
> > (24dBi antennas at both ends, 1Watt amps, LNA).  Of course, that would
> > require the cooperation of other hams, but, overall, I think I rather
> > trust
> > the cooperation of hams than depend on a board of directors whose
interest
> > is maximizing shareholder revenue. You might even be able to beat the
HOA
> > with this... I don't know what it takes to get a MMDS license, but once
> > you
>
> Luck and a lot of money!  They are awarded 4 channels per area by the FCC
on
> a lottery basis.  You are probably looking at a lot of money.  There is a
> non refundable $155 filing, or application fee and time limits for
> establishing a working system.  System price is probably on the order of
a
> hundred grand.  I'm certain that could be done cheaper, but there is a
> mountain of paper work and consulting fees prior to submitting the
> application.

But hams work for free, so, presuming that mountain of paperwork is within
the capability of a suitable group of hams, that's not the sticky point.
Likewise, $155 is a trivial expense in the overall context of things. The
hardware cost (if there's some FCC requirement for a particular level of
provisioning) could be the hangup.  And, of course, the lottery.

>
> > establish that you NEED that 100 ft tower to get your high speed
internet
> > access....
>
> I could put in the equipment for line-of-sight wireless to the
*relatively*
> local ISP and only have to pay for the dedicated IP address over my web
> hosting and a connect charge, but I'm not sure just what equipment to use.

I'm not sure that a 802.11 type WISP is covered under the OTARD rules.  MMDS
certainly is.

For the "backbone" kinds of things, OTARD is immaterial, as is licensing to
try and beat the HOA.  Actually, though, I think they're thinking about
unlicensed U-NII so they don't have to get wrapped around the axle about
things like "no compensation" and "no business".  You don't want to have to
agonize about whether getting your work email via the wireless backbone fits
within the description of amateur radio.
They're point to point links for the most part, so this is realistic.  The
"mountaintop to user" link, depending on the mode, would be amateur (i.e. 2m
voice links)

>
> 802.11g is 56 or 58K, but that is full duplex which makes it close to 100
> base T (which is not) without as much problem with collisions OTOH I don't
> think they use it for that kind of distance.
I think you mean 54 Mbps (for the payload... the sync sequence is lower)),
and with overhead, the "throughput" is somewhat lower.

Indeed, 802.11g isn't used for more than 100m, however, you can leverage off
the readily available equipment.  Much cheaper to use mass-market and adapt
to that paradigm than to roll your own microwave links. (FWIW, 100BaseT can
be full duplex. A lot depends on the switches in use, and whether your NIC
driver supports it.. There are two pairs, one for Tx and one for Tx.
Passive hubs force half duplex, and some NICs (drivers really) come up in
half duplex mode, but these days, almost every switch supports full duplex
on 100BT, and so does the NIC)

> What do they use for say 5 to 10 miles without having to go to a high
priced
> system.

I know of several people using essentially unmodified 802.11b hardware with
an external high gain antenna for this kind of distance.