Hacker News Reader

Comment by rkagerer | Hacker News Reader

Can someone explain this excerpt to me?

NATO is currently investigating future internet backup routes through satellites in the case of undersea cable failures. But that technology is only in a preliminary, proof-of-concept stage and may be many years from real-world relevance.

We already have satellite links and can configure our networks to route through them when a primary route disappears. Why are they saying this is "future tech"?

Are they talking about backups at a bulk backbone level for which we simply don't have enough satellite capacity?

View on HN

These cables offer enough bandwidth to service an entire country and then some, because they're also passing along data coming from further on. We have no satellites that can offer anything close to that rate.

IIRC starlink might try

Starlink is the highest capacity satellite network by far and the entire constellation’s RF capacity would struggle to compete with a single undersea fiber cable.

It’s hard to put into perspective how impressive modern fiber is. Each wavelength carries many gigabits per second so FDM lets you pack terabits/sec onto a single fiber. Undersea cables can easily bundle 50 fibers, with the primary constraint being the repeaters. Confluence-1 is supposed to carry 500 tbps with just 24 pairs.

In comparison, at a generous 50gbps a sat, that’s 1000 satellites of capacity in a single cable. And that’s not even possible logistically due to orbit distributions and inter-sat interference aiming all of their RF beams at one spot.

All that said, starlink is an incredible technology for “last mile access” that will globally eat the market for remote locations with nothing coming close to competing. It’s just that undersea cable is still moving orders of magnitude more volume.

Ok that makes sense, and jives with what I thought.

Is there any proof of concept in the works that could radically change the equation?

Probably challenging to sufficiently harden the links and ground equipment, and to guarantee QoS for military applications

Capacity is the issue. Undersea cables typically have terrabits per second of capacity. Newer cables are hundreds of tbps.

Satellites just don't have that kind of capacity at this point. Certainly, you can aggregate several satellites, and you may get to enough capacity for your needs, but it will be a lot less capacity than a cable and you may need to have significant usage restrictions in one form or another.

> Why are they saying this is "future tech"?

I read it like "we're doing proofs of concepts but it'll take some years before it's production ready", not some wishy washy "future tech". It's in the pipeline, just that the pipeline takes some time.

I expect Starlink and other satellite providers can do QOS and prioritise critical traffic so it wouldn't have to compete with the 99.99% of streaming services in an emergency.. But I may be wrong.

Reply on news.ycombinator.com