Improved Decoy-state and Flag-state Squashing Methods

Abstract: In this work, we present an improved analysis for decoy-state methods, enhancing both achievable key rates and recovering analytical results for the single intensity scenario. Our primary focus is improving the shortcomings observed in current decoy-state methods, particularly recovering results when employing no decoy intensities. Our methods enable the continuous interpolation across varying numbers of intensity settings. Additionally, we extend decoy-state techniques to encompass scenarios where intensities vary depending on the signal state, thereby relaxing the constraints on experimental implementations. Our findings demonstrate that a minimum of two intensities are sufficient for high asymptotic secret key rates, thereby further softening experimental requirements. Additionally, we address inherent imperfections within detection setups like imperfect beamsplitters. We derive provable secure lower bounds on the subspace population estimation, which is required for certain squashing methods such as the flag-state squasher. These analytical bounds allow us to encompass arbitrary passive linear optical setups, and together with intensities varying with each signal state, lets us include a broad class of experimental setups.