Given how astounded I still am that, 30 years on from hugely expensive (and thick) ethernet cables, taps etc that gave us 10megabit/s over a max of (was it 150 metre? I forget now. Not very far) we can do 20megabit/s over a kilometre or two on simple twisted pair phone wire that really wasn't designed for it...
What's next on the horizon?
There's certainly room for improvement in IP over satellite services.
Geostationary communications satellites, or "
extra-terrestrial rocket station radio relays" as they once were known, were foreseen by Arthur C. Clarke right back in 1945.
These satellites hover at an altitude of some 30,000 kilometres above the tropics. Fixed in what we now call the "Clarke Belt". An orbital belt named in honour of
the late great science fiction writer.
There is an enormous propagation distance - 60,000km - to the Clarke Belt and back again. And that brings with it the major problem of signal latency. Sending/ack'ing a TCP packet via the Clarke Belt adds a round-trip latency of some 500 milliseconds. A half a second delay to the communication. [For proof, recall that photons travels in free space at ~300,000 kilometres per second]
So this currently precludes the use of IP-over-Sat for any real-time packet-based communication where data integrity is critical. A damned nuisance!
More recently, however, NASA has gotten its International Space Station (ISS) to hover at just 300km altitude. An amazingly low altitude. The same trifling distance as London to Liverpool. In fact, 300km is so low that the ISS should be almost visible with the naked eye. Certainly it should be obvious on a clear night with
decent optics.
So what does this mean for IP-over-Sat? Well if a "transponder" (or two) was somehow glued to the side of the ISS, and if the ISS was fixed to a geostationary point at a 300km altitude, then we could have IP-over-sat, but without that previously prohibitive latency of 500ms.
The round-trip delay for an ultra low-altitude satellite link between NY to London would be almost the same as for a fixed wire connection. No need to lay any more of those costly transatlantic submarine cables!
So that's perhaps the vision for broadband in the 21st century. Farewell to DSL and hullo to IP-over-Sat!
How long before we see satellites hovering at perilously low altitudes? Mind your heads! Perhaps no higher than the F2 layer of the ionosphere (~275km)? And offering terabit bandwidths at microwave frequencies with minimal latency?
If ever that becomes reality, we can finally kiss goodbye to the likes of BT and Virgin Media.
cheers, a
