All this news has got me thinking… how would you go about making a communications line that was mechanically very difficult to tap?
Filling the sheathing with pressurized gas has already been tried, and bypassed by e.g adversaries working in pressurized caisson-style rooms that equalized the pressure.
Here’s another idea.
One trick the TSCM folks love for finding taps on phone lines is a time domain reflectometer, or “cable radar.” It sends a pulse down the line, and the reflection profile produces a graph of relative cable impedance changes as a function of distance. Conventional taps always produce significant impedance changes, even “inductive” taps that can’t be detected with simple voltage and current measurements. However, there are taps out there that can defeat the TDR (because they perturb the E & H fields so little there’s no significant impedance change).
I’d wager that penetrating a coaxial cable, however, would always produce a significant impedance change. Even if you used a non-conductive and non-inductive drill, the presence of added dielectric between center and sheath ought to change the impedance at that location significantly.
As it happens, coaxial cable with a tubular metal core is easily available, Granted, it’s thick and stiff as hell, but the impedance is very precise and stable.
Run a fiber-optic line down the center of such a tubular coaxial cable — and connect the coax itself to a constantly-running TDR unit — might produce a strongly tap-resistant line. The instant someone tries to penetrate the coax (necessary in order to tap a fiber line, I think) you see on the TDR exactly how many feet away someone’s messing with your cable.
For jam-resistance and antispoofing, the TDR would need some kind of pseudorandom sequence for each pulse (commonly used in military radars, see the “Russian Woodpecker” for a well-studied example).