Cheap TDR for fibers?

[Thomas Shaddack]

The laser diodes used in eg. CD players have a feedback photodiode, 
sensing the laser's optical output.

If the lasers used for optical fibers have similar mechanism too, and if 
the diode is sensitive to the light coming to it not only from the chip 
but also from the fiber itself, and can react quickly enough with high 
enough sensitivity, maybe it could be exploited.

In chosen moments, we could then send a short pulse of laser light into 
the fiber, then watch the signal from the feedback diode, what gets 
reflected back from nonhomogenities on the fiber. This would give us the 
distances of all the splices and connectors, and let us know immediately 
(if the test is performed eg. once per 5 seconds or with similar short 
period) that there is an attempt to compromise the line underway. 
Comparison of snapshots from longer periods apart could also serve to find 
deterioration of the signal path before it results in failure.

The advantage of this approach, if possible, is the ability to add the 
functionality without having to modify the optical transceivers 
themselves.


It sounds too good to be true, so it probably won't work, but I may be 
wrong...