Value of spatially multiplexed SNSPDs with photon-number-resolving capabilities
We develop a combinatorial and Bayesian framework to quantify the information gained by increasing the local photon-number resolution of multiplexed detection systems based on superconducting nanowire single-photon detectors (SNSPDs). Motivated by recent studies that attribute photon-number resolving capabilities to a single SNSPD overcoming the typical on/off detection mechanism, we generalize the existing formulation for the binary devices by deriving the probability response tensor of multiplexed systems based on SNSPDs with an arbitrary number of resolvable states. We include finite efficiency through binomial loss convolution and, separately, assess the impact of dark counts. From this response tensor we derive a generalized N-photon detector fidelity, a Bayesian photon-number reconstruction, and the corresponding reconstruction fidelity. We particularize this formulation to study 2-, 3- and 4-state multiplexed systems with different numbers of detectors and show that increasing local photon-number resolution systematically improves both detector fidelity and Bayesian photon-number reconstruction. This substantially relaxes the quadratic scaling between detector number and resolvable photon-number required by purely binary multiplexed arrays. These results provide a quantitative value-of-information assessment for multi-state photon-number-resolving detector arrays.