Suppose we are attenuating up to 0.2 mean photon number. So, every RF cycle will not contain a photon. So while providing a 1-bit delay, it is possible that the interference can happen between a vacuum state and a photon or a vacuum-vacuum state or photon-photon state.
If there are two neighboring photons, then we will get the exact phase difference after interference resulting in the correct bit assignment. What will happen if a vacuum-photon state arrives at the other end of DLI? Will they interfere or simply pass with a 50% chance of detection in both the detector? This will generate errors in the key string. How to rectify this error?
I want to do sifting but got stuck at this point.