how do transcontinental lines work?

[rc240sx]

If we have an internet connection from new york to california, how does a signal travel that far? Even multimode fiber can only go 2000 meters unrepeated. Are there many repeaters along the way, many COs? How does it span for 3000 miles?

 

[Lord Evermore]

They go through central office equipment regularly. There's actually some extra latency induced due to that, even though with fibre the light would be moving somewhere around 80% of the vacuum light speed. They work pretty much the same as any phone line that has to reach that distance. Depending on what it is, it may be a simple repeater, or it may go through an aggregation system.

Of course, that just applies to a backbone line, or a telco trunk line. An actual user's Internet access just gets routed through network equipment with multiple connections until it reaches one of those backbones.

 

[ScottMac]

Transcontenental is nuthin' - 3000 miles, over / under land, plenty of power, plenty of room for repeaters ... no sweat.

Think about Trans-Pacific; 7000 miles, under water, no local power ... there's some pretty neat sh!t going on there.

Some of the newer runs have "light amplifiers" (boosts the light without first converting it to an electrical signal), special dopeing of the glass, high-intensity lasers ...

Cool stuff for just transmission medium.

FWIW

Scott

 

[Fencer128]

They could use single mode fibre (to manage the dispersion) together with Erbium doped fibre amplifiers/repeaters - as they do in long haul systems.

The repeaters would have to be every ~100km (it used to be less but I think technology's improved).

It's off the shelf tech.

Cheers,

Andy

 

[cmetz]

rc240sx, from NY to CA will go through multiple land-based relay centers and amplifiers. It is incredibly unlikely that you have one fiber from city to city, they're going to be long bundles with patches between at the relay centers.

Fiber's main advantage over copper is its ability to carry signals so much farther between repeaters. And on top of that, there are Erbium Doped Fiber Amplifiers, which are an all-optical amplifier that is built into long-haul fiber itself (a second laser is used to provide the excitation energy for the EDFAs, they then absorb signal laser and re-emit at higher amplitude). Trans-oceanic fiber really is a straight fiber for thousands of miles, because it's kinda hard and kinda pointless to put a relay center in the middle of a major ocean.

But back to your question: if you had a dark fiber from NY to CA, then it would be a bunch of fibers from various bundles with patches in between. There probably wouldn't be that many patches, though, because each one has a loss cost. Much more likely, however, is that you have a SONET link from NY to CA, in which case there is likely to be active equipment in between - maybe some add-drop multiplexers, probably some WDM gear, and definitely patches and SONET repeaters. Most carriers' long-line folks are not terribly open about what exactly is in the middle of your link because they might not be sure themselves (these networks are pretty arcane sometimes) and because they want the flexibility to re-route or otherwise change things, hopefully for the better. The key is that you have an circuit (e.g., SONET) that you can shove bits in one end of, and delta-T later, they come out the other end. For most users, that's all you need to know. (the folks who are dealing with things so critical that they need to know exactly what's in the middle often own their own fiber anyway)