[MLB-WIRELESS] Applications on the melb-wireless network

[Ben Anderson]

>   Ben> At one stage canada had a grand government plan to lay a
>   Ben> massive backbone across country and basically lay it open for
>   Ben> all to use (they overengineered it something chronic).  I
>   Ben> haven't heard of it for ages now, I wonder if it actually went
>   Ben> anywhere...
>
> Canarie (www.canarie.ca) by any chance?  although i thought this was
> only for educational use?

I didn't see anything limiting it to educational...  though even, it shows
what you can do with only 120million...


>   Ben> The telco's stand to make **more profit** this way, not less.
>   Ben> And if they run the numbers, I'm sure they'll stand back and
>   Ben> let someone else worry about laying cable (price is high,
>   Ben> uptake is low, and therefore they don't get the return on
>   Ben> investment that's expected).
>
> well, i'm not an economist, and so my comments must be taken in light
> of that fact.  but ... i've always felt that there was more ideology
> than economics in the methods of privatisation of telecommunications.

More politics, I'd say.  You allow us to sell this, we'll fund more greenies
stuff...  (it went something similar to that IIRC)


>   Ben> Passing data back and forward for someone else is considered
>   Ben> "carriage" of the data, and requires a carrier licence (from
>   Ben> some of the legal info I've read today)...  Bye bye public
>   Ben> access points, bye bye mesh.
>
> this is where it becomes adventageous for the wireless community to
> formally establish itself as a legal entity.  then the data is not
> being carried for someone else, it's being carried for the "entity".
> there's been some suggestion that a standard "club" entity would be
> suitable.
>
> intra-entity carriage is legal, i believe, without a carrier licence.

Perhaps...  it depends how you interpret the law...  passing information for
someone elses behalf is still being done inside the entity, by other legal
entities.  If this were a way around it, why not just start an entity called
"the internet" and be done with it?
This should be dealt with by a laywer, not hacks like us (I assume you're
not a lawyer).  I want to believe it's legal, i think it makes sense to be
legal.  But still, every lawyers advice I've seen about this stuff basically
reccomends against it in some way or another.


>   Ben> if it's ubiquitous, and the infrastructure works, then they
>   Ben> won't have to compete.... they simply provide services to the
>   Ben> network, and they no longer have to spend big money setting up
>   Ben> the network.  Costs go down, income stays similar, profits go
>   Ben> up.  I don't think they'll complain.  Or is there something I
>   Ben> haven't considered?
>
> you've forgotten the millions (and in some cases billions) of dollars
> in debt that they have to service which they spent on cable and are
> about to spend on 3G.

Which is crazy, that amount of debt...
But still, I maintain my stance.  If the carriers no longer have to support,
or roll out further infrastructure, just plug into existing bandwidth to
provide the service, it's just cost them a whole lot less to do the business
that they're actually making profit on (!).  Rolling out and manging
broadband at the moment is a big money-sink, gambling on the future type
affair.
Perhaps the infrastructure could be purchased as part of the UPN.

> it's not the fact that they don't have to spend the money for future
> infrastructure, it's that they cannot afford to have consumers using
> cheaper services because they no longer get the profit margins
> required to service their debt and the stockmarket.

Not about cheaper services.  Read my above clarification :)

>   Ben> One of the plans I've gotten told to get my head outta my arse
>   Ben> for is the flying access point.
>
> burt rutan (the guy who designed the gossamer albatross and the other
> pedal/solar powered planes in the 80's) has a company that is building
> a plane designed for this sort of thing.  i believe that someone is
> planning to do this over Chicago (?) using three shifts of 8 hours.
>
> see  http://angeltechnologies.com/

Sweet, burt rutan rocks...  Definatly deserves some kind of living legend
status IMO.

Though that's one expensive looking jet aircraft he's got...  A chunk of
helium/hydrogen is a lot cheaper to float than a heavier than air jet
aircraft.  And a balloon type technology is feasable for hacks to build in
their backyards, at the moment just the jet engines for such a heavier than
air device is well beyond what most of us would willingly spend on such a
project.

> i've also heard rumours that some airlines are planning to look at
> this: over the continental US and Europe, there are a huge number of
> planes in the air at any one time.  if they ran a switch and a
> transceiver dish each with some smart handover software, they'd be able
> to provide a 24/7 coverage of most populated areas.
>
> but this is getting a little off-topic ...

Not really.  Still wireless networks.  Exploring wireless options for
melbourne.  Melbourne wireless.  There's a link to "topic"  :)


>   Ben> Another consideration is the 802.11a stuff, with good line of
>   Ben> sight (no fresnel zones to worry about, etc) pretty reasonable
>   Ben> sized antennas should make 54Mbit access reasonable.
>
> with Proxim's X2 stuff, you get 108Mbits for the same card cost.  so
> you can double these numbers.

Not really.  there's 13 independant 54Mbit channels.  Pairing two channels
together for 108 doesn't allow me to multiply the 702Mbits, as now there's
only 6.5 108Mbit channels.
Still, 7/10ths of a gigabit would make a shit-hot UPN for most traffic
(though some limiting metric would still likely need to be imposed as the
network becomes more and more loaded)
My calculations ignored the need to bounce signals off the AP, effectilvy
halfing the numbers I gave.  So 351Mbits...  Still nothing to sneeze at,
much better than an effective 300kbits bouncing two ways off a 802.11b AP.

>   Ben> And multiple devices (up to 13) for 13 independant 54mbit
>   Ben> channels... 702Mbits. That could probably scale to 2000 users
>   Ben> with very good access, perhaps 5000 with acceptable, and nearly
>   Ben> 13,000 with 56k+ access (though average use would dictate the
>   Ben> number could probably scale to about 5 or 6 times this, and
>   Ben> more if we could put up with congestion during peak times)
>
> this is 5GHz gear -- does it have similar properties to 2.4GHz ?  or
> is it more subject to weather interference, for example?  i know the
> range that they quote is roughly half that of 802.11b -- why is that?

They need a necessarily higher signal-noise ratio, for the greater
bandwidth.  And there's additionally more line of sight problems.  Line of
sight isn't a massive issue with a flying antenna ;)

>   Ben> Would you like to comment on feasability issues you see with
>   Ben> the balloon idea?
>
> the issue with satellite services is latency -- fibre is sooo much
> quicker.  do people recall the old 700ms satellite hop to the US?

Yup, it sucked muchly.  The southern cross fibre network has that down to
70ms.  Nice enough to play quake competitivly :)


> what sort of latency do we introduce at 10km up?

Assume roughly the speed of light travel time, 300,000km/sec.
About 0.033msec.  Double it for the bounce-back. And still, the processing
time in the AP far outweighs the transmission time.

> other than that ... i think it's a fine idea.  a few solar cells, a
> newish laptop, a bunch of pringle cantennae, and the worlds biggest
> wine cask bladder and we're off ... ! ;-)

I'd throw the laptop out, use a leightweight access point.  I'd use a
30degree (or so) patch antenna (flat, low wind resistance), or a 'flat dish'
(probably a much better choice) with a fairly wide spread pattern for a
dish.  There's control systems, gps, servos and propellors to be able to
control.  There's a compressed can of helium or hydrogen and solenoids for
altitude control.
Remember, a displaced litre of air lifts about a gram.  The batteries to
last a night are going to weigh the most (fuel cells would be good, but
the're largely still unavailable.  Beauty of ths would allow the hydrogen
cylinder to provide both altitude and power).  2kg I'd suggest as a useful
maximum for a still maneuverable craft in reasonable winds (would require ~
2m^3 balloon, probably a fair bit larger, as I haven't compensated for lower
displaced air density at higher altitudes in my calculations).

is there anything I've missed?

Ben.



--
To unsubscribe, send mail to minordomo at wireless.org.au with a subject of 'unsubscribe melbwireless'  
Archive at: http://www.wireless.org.au/cgi-bin/minorweb.pl?A=LIST&L=melbwireless
IRC at: au.austnet.org #melb-wireless