Richard Clark wrote:
Hi Ian,

This is called "conflict of interest," which discounts those same
lecturers' and professors' credentials.

Well, I for one am not going to instantly take the view that my
lecturers' credentials are not solid. Of course they're going to put
positive spin on it, mainly because for their part (physical layer and
comms protocol stuff) they've been successful so far. Trials in Sweden
were a total success, albeit on a smaller scale.

I'm still not convinced that the idea will ever actually be realised,
but nevertheless it's still a very interesting one!

You've missed the point Roy made. Adding connections (more HAPs) does
not add more bandwidth. Those extra HAPs will be competing for the
same (now diminishing by proportion) spectrum.

That'd be why there is ongoing research into the various multiplexing
techniques so many users can use the same piece of spectrum and not
cause *too much* interference with each other. Using spreading codes etc
the other signals just appear as a little bit of extra background noise
so I am led to believe. (I will get the full story on this sort of stuff
in the next academic year).

Surely though even for broadband internet home users will not need to
exceed 10Mbps speeds, what would be the point? Who needs to get a web
page served a second faster, bearing in mind bandwidth limits at the
server end as well as the end user connection. The 3rd generation
mobiles were not looking to exceed 5Mbps per handset (which is a hell of
a lot of data) at the very most and that is more than capable of
streaming video etc (albeit at lower resolutions for the handsets).

The bandwidth requirements for cellular voice calls is minimal in
comparison to data requirements. A phone line is only 64kbps, and
cellular (gsm) data rates are less than that even and still provide good
(enough) voice reproduction!

When there's existing hardware (after all, no one is telling the
consumers to throw away their phones and buy HAP versions), and
Hindenberg technology is a century old; then any proviso "there is
still a lot of work to be done" translates into SEND MORE MONEY - a
message tape with an infinite loop.

Basically as I understand it they'd be looking to use current Wi-Fi,
Wi-Max and GSM technologies etc so why would there be a requirement to
change hardware?

The only thing that would need changing with the broadband data
downlinks to serve internet would be gateways with directional antennae
to serve buildings etc.

Ask researcher1: "can I float a balloon?"
researcher1: "Sure, no problem."

Ask researcher2: "can I transmit and receive from a height?"
researcher2: "Sure, no problem."

Ask researcher3: "can I find a stabilizing platform?"
researcher3: "Sure, no problem."

Ask researcher4: "can more connections serve more customers?"
researcher4: "Sure, no problem."

The sum is not equal to the whole:
Ask customers: "can you still hear me?"
customers: "What the ****! My line is dead."

This is the same with any new technology! Just look at 3G services in
the UK, it's taken them a while to get network coverage any where near
comparable to the already existing 2 and 2.5G network infrastructure.

Do you suggest that we just give up carrying out research into this sort
of thing? Maybe we should have stuck with the original optical telegraph
rather than develop methods of signalling using electricity...

I have no personal/pecuniary connections with this project, as I'm only
an undergrad student, but I think to dismiss it out of hand as a non
starter is a bit harsh. It does have the potential to work, whether it
ever gets deployed is another matter...

--
73, Iain M0PCB/P