They are talking about laying a fiber optic cable from Japan to Europe

[Winterblues]

It will go under the Arctic Ocean at a cost of one billion dollars. With the ice having shrunk to a point where through passage is now a reality, it has become feasable. The project will begin either later this year or early next year and be completed by 2013. Heard this on morning news. My question is with everyone using satellite communications more and more, what is the point? I guess if we were to have a major solar flare and lost satellite communication, the world would still be connected. :shrug: Sounds like an interesting project anyway.

[DisgustipatedinCA]

1. latency

Using a satellite introduces latency in the round trip that the IP packet has to make. This slows communications. Some programs are very time sensitive (voice over IP, for example) and would not tolerate the latency added with satellite communications.

[muriel_volestrangler]

2. No satellite delay is one reason

and bigger bandwidth.

[KharmaTrain]

3. Satellites Are Vulnerable...

A cable can operate for decades while a satellite's "shelf life" is less than 10 years and are costly to launch. Also their communications can be intercepted where a "landline" is far more secure. The cable has been feasible for years...and the shrinking icecap has little affect as the lines would go under the pole just like our submarines do.

[Winterblues]

4. I did not realize submarines were able to lay cable.

It seems to me that a lot of cable has to be on some platform before it actually gets submerged and I don't believe submarines are large enough to carry that much cable. No it is important that the ice has receded or it would not be feasable at all.

[KharmaTrain]

9. Submersibles...

Yes, a platform ship is required for this kind of job, but much of the heavy lifting are done by submersibles. Recently NatGeo had a program about how they are able to lie pipeline and cable under long stretches of the oceans and at greater depths. Also holes can be drilled into the ice cap to assist those machines...with minimal impact to the environment (not like building an oil platform).

No doubt that the shrinking of the ice caps are a factor in building this cable but I don't see this creating any long lasting damage to the ecosystem that doesn't already exist. Who knows...may even speed up internet connections.

[DBoon]

5. You are right about interception

My first thought is they are doing this to avoid having traffic route through the US, and intercepted by the NSA

[Vincardog]

6. Bingo we have a winner "to avoid having traffic route through the US, and intercepted by the NSA"

[sofa king]

8. That's not a problem for NSA.

Remember all those fiber lines that were cut during the Bush years?

http://www.wired.com/science/discoveries/news/2005/09/68894

All they had to do was rip off one American inventor to be able to read everything going through those lines that were "accidentally" cut.

[Winterblues]

7. Except it is supposed to touch ground in Alaska to provide access to rural communities

Like Nome and Pt Barrow and interior Alaska, then it will go back under water to Greenland. Then from Greenland on to Europe.

[hunter]

11. One of the failures of U.S. intelligence is this absurd faith in "secrets."

We're playing poker in a world where everyone else is playing chess.

In these modern games skulking about to peak at the other guy's cards gains you much less than it used to. Most of the action is taking place out in the open, on the chess boards.

[Statistical]

10. Sat costs a magnitude (like 100x) more than fiber on bandwidth basis.

99.9% of digital traffic flows of fiber connections. The share of two-way traffic over sat connections (excluding broadcast traffic) is very small and dropping every year. Sat will never even come into same ballpark as cost parity with fiber. Price of fiber falls by about 70% per decade (same fiber more bandwidth using better lasers) while cost of sats is roughly static.

http://www.satsig.net/ivsatcos.htm

Digital communication sat costs in the $300 mil range to build, test, launch, insure, and maintain. It will likely last 20 years or so.

The break even cost then is about $80,000 per year for a 1Mbps downlink.

That's right 1 Mbps (roughly 1/5th to 1/10th you cable internet connection speed).

Now that 1 Mbps is dedicated but it still runs about $80,000 a year for a mere $80,000.
In comparison $80,000 will buy you about 120Mbps dedicated from New York to London by fiber.
1Mbps vs 120 Mbps. Same service level, same dedicated (unshared) connection, same cost. Roughly 100x as much bandwidth.