Advancing Gaussian Boson Sampling with Realistic Photon-Number-Resolving Detectors

Abstract

Gaussian boson sampling (GBS) represents a compelling near-term quantum computational model that harnesses nonclassical Gaussian states processed through a linear optical interferometer. While GBS holds significant potential for demonstrating quantum computational advantage, the certification of such experiments remains an active challenge. Validation techniques are often constrained by the type of detectors used. The reliance on ideal photon-numberresolving (PNR) detectors, which can distinguish precise photon counts, limits the practical applicability of these methods, as most real-world detectors exhibit non-ideal characteristics, including dead time and limited discrimination capability. In this presentation, we explore the theoretical framework necessary to incorporate realistic PNR detectors into GBS experiments. We derive a general photocounting probability distribution, expressed in terms of functionals like Hafnian and Torontonian for different detector models, to better match practical conditions. This result extends existing validation methods, enabling robust tests for GBS setups with imperfect photon detection. Our framework not only enhances the interpretability of experimental outcomes but also refines the boundaries of classical simulability under realistic constraints. Additionally, we discuss how imperfect detection affects the outputs of GBS, noting implications for applications in graph-based problems and quantum photonic systems. By addressing the gap between idealized models and experimental limitations, our results contribute to the design of more resilient quantum optical experiments. This presentation will provide an accessible overview of our approach, linking foundational concepts to experimental considerations. The presented work aligns with ongoing efforts to bridge theoretical advances and experimental realizations in photonic quantum computing. Our results have been detailed in the article [1].