Performance assessment of SPAD arrays for coincidence detection in quantum-enhanced imaging

Single-Photon Avalanche-Diode (SPAD) arrays find extensive use in quantum imaging techniques that exploit entangled-photons states to overcome sensitivity limitations of classical imaging. Thanks to their compactness, low-voltage operation, single-photon sensitivity, absence of readout noise, and high frame-rate, SPAD arrays are particularly suited to detect temporally correlated photons over a scattered background. This work presents a scheme useful to model a generic quantum imaging measurement set-up, with its losses and non-idealities, and it provides the resulting calculations of pair rate (in case of quantum states made of two photons) and spurious single-photon rate at detector level. The computed rates are used to evaluate the performance in terms of signal-to-noise ratio of a possible SPAD array architecture based on an on-chip photon coincidences detection, followed by an event-driven readout, which transfers only the addresses of those pixels involved in the coincidence event. Although bringing plenty of advantages in terms of power consumption, data storage, and readout time, especially as the pixels number increases, the intrinsic non-ideal operation timings of this architecture are linked to three possible cases of wrong detection. A detailed computation of these error probabilities is provided, together with a discussion about which design parameters most influence the detected signal quality. Since every on-chip coincidence detection and event-driven architecture is characterized by those same finite operation timings, the presented computation method can be considered a useful tool to optimize the design of detection systems used in quantum imaging and microscopy framework.