You used to be able to get diffraction gratings that could measure the temperature of a bulb filament accurate to 10 degrees or so, I think they were called Bragg gratings. They were film/glass with lines accurately spaced. Using two of these you could identify the wavelength of the light...
Ignoring reflections and assuming a parallell tunnel and also assuming a direct line of site between transmitter and receiver, then the strength calculations are exactly the same as in free space. Just simply calculate the received strength as a function of 100 + x metres. If you include...
Although waveguide method will work in theory, it is an unreliable route for this problem.
Perhaps the most reliable route is to run a length of leaky coax from a base station throughout the length of the pipe and transmit to and receive from this cable using low cost short range transmitters...