Can optical fibers transmit in both directions simultaneously? Both single mode fibers and multi mode? What about if they we choose non-interfering wavelengths?
5 Answers
Theoretically, such concepts are good to be imagined and, in fact, it is possible to make one that can communicate bidirectionally, but for reliability it is preferable to have a duplex with two fibers.
-
\$\begingroup\$ Concise answer. I tweaked the punctuation and grammar a bit to make read nicer. \$\endgroup\$– user65586Commented Dec 7, 2015 at 14:45
Yes, it's done all the time; not "just in theory" and it's not at all "better to have two fibers" - in long-distance networks where fibers are expensive, many different wavelenghts travelling on both directions are a common, everyday state of affairs. The equipment to get them injected and separated at each end is expensive, so that sort of thing is not seen where fibers are cheap to install. Look up CWDM and DWDM (Coarse Wave Division Multiplexing and Dense Wave Division Multiplexing) for more detail.
On my network I have one set of 1310/1490 BiDi SFPs; I'd have more if I could find them cheaply enough (though they are MUCH, much cheaper than the devices above) - on the particular link that set is installed on they allow 3 links to run over 5 fibers where 7 were damaged, and only 2 links were possible with the dual-fiber SFPs (still in use on 2 pairs) - they work perfectly and only need the single fiber.
Light travelling in both directions through a fiber is no problem. Paths of in- and outgoing light can be seperated/combined by beam splitters.
Interference happens only if two different beams hit the detector. Then both beams may interfere (coherence required).
In fiberoptic gyroscopes this effect (Sagnac effect) is even wanted.
Apparently that is possible and there is at least one commercial product.
I haven't seen them used in practice, though.
Photons don't interfere with each other so in theory on a perfect fiber it would be possible to use the same wavelength in both directions but in practice that doesn't really work because of reflections.
On the other hand different wavelengths in different directions is very much possible.
You can easilly get transcivers that use two widely spaced wavelengths on the same fiber for the two directions. The wide spacing of the wavelengths is presumablly done to keep the splitting optics cheap. These are commonly used in fiber to the home type applications on single mode fiber though appparently it is also possible to use them on multimode.
There are also systems that can put many different wavelengths on a fiber. This is known as "wavelength divison multiplexing". Due to cost such systems are usually only used on long distance backhaul links. Most commonly all the wavelengths on a given fiber are run in the same direction but if the number of fibers to a given destination is small there can be advantages to running both directions on the same fiber.
-
\$\begingroup\$ Photons don't interfere with each other Can you justify this statement? Photons can share local energy states due to their bosonic nature. But any coherent narrow band light beam will actively interfere with itself (due to propogation broadening) and other coherent light sharing the same conduction and detection channels. \$\endgroup\$– crasicCommented Dec 7, 2015 at 18:32