Robustness and optimization of N00N-state interferometry Preprint: APS/123-QED
Romain Dalidet 1 , Anthony Martin 1 , Louis Bellando 1 , Mathieu Bellec 1 , Nicolas Fabre 2 , Sébastien Tanzilli 1 , Laurent Labonté 1 · 2026
Read Robustness and optimization of N00N-state interferometry Preprint: APS/123-QED by Romain Dalidet 1 , Anthony Martin 1 , Louis Bellando 1 , Mathieu Bellec 1 , Nicolas Fabre 2 , Sébastien Tanzilli 1 , Laurent Labonté 1 (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-13144.txt.
Introduction
Romain Dalidet and co-authors map the trade-offs that set real performance in N00N-state interferometry. Ideal scalings ignore loss, state imbalance, and photon number, which all interact in practice. In a folded Franson setup with partly entangled states, they derive exact fringe visibility and Fisher information. Loss asymmetry can always be offset by input imbalance to restore full contrast. Fisher information peaks elsewhere, though, which marks a true trade-off. They also find the least entanglement that keeps a quantum edge.
Catalog record
| CREATOR | Romain Dalidet 1 , Anthony Martin 1 , Louis Bellando 1 , Mathieu Bellec 1 , Nicolas Fabre 2 , Sébastien Tanzilli 1 , Laurent Labonté 1 |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 31523 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.13144
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Robustness and optimization of N00N-state interferometry Preprint: APS/123-QED Romain Dalidet 1 , Anthony Martin 1 , Louis Bellando 1 , Mathieu Bellec 1 , Nicolas Fabre 2 , Sébastien Tanzilli 1 , Laurent Labonté 1 Abstract Quantum-enhanced interferometry is often discussed in terms of ideal resources and asymptotic scalings, whereas in practice its performance is set by a delicate interplay between losses, state imbalance, and photon number. We address this interplay in a folded Franson interferometer fed with partially entangled N00N states, treating asymmetric losses and tunable input imbalance on equal footing. From exact detection probabilities we obtain closed-form expressions for the fringe visibility and the Fisher information, and show that these two figures of merit respond very differently to imperfections. In particular, we demonstrate that perfect interference contrast can always be recovered by compensating loss asymmetry with an appropriate input imbalance, while the Fisher information generally peaks at a distinct operating point, reflecting the irreducible trade-off between coherence restoration and signal attenuation. By determining the exact optima and benchmarking against single-photon strategies, we identify the critical loss and minimum entanglement required to maintain a genuine quantum advantage over optimized single-photon strategies under identical loss conditions, and establish their scaling with the photon number N N . Beyond delineating the fundamental trade-offs between loss, entanglement, and sensitivity, this work establishes a comprehensive theoretical framework that both underpins and extends the experimental demonstration of quantum advantage reported in dalidet_QA , providing a unified description of the relevant operating regimes. I Introduction Quantum-enhanced interferometry is often presented as a direct route beyond the standard quantum limit (SQL): prepare a nonclassical probe, imprint a phase, and read out an interference fringe with an increased slope Barbieri2022 ; Paris2009 . In realistic experiments, however, the achievable performance is not determined by a single imperfection but by a coupled landscape of experimental parameters in which losses, state imbalance, interferometer asymmetries, and photon number jointly shape the available information Lee2002Rosetta ; Demkowicz2015 . In this landscape, some parameters are genuinely critical, as small deviations can abruptly suppress any quantum advantage Rubin2007 ; Oh2017 , whereas others can be tuned to partially compensate each other Lee2009Loss ; AtamanMishra2024 . A central question for practical quantum sensors is therefore not whether an ideal N00N state can beat the SQL Mitchell2004 ; Nagata2007 ; Afek2010 , but rather how quantum advantage deforms under realistic imperfections and which experimental degrees of freedom can be exploited to restore it ThomasPeter2011 ; Qin2023 ; Nielsen2023 . This issue is particularly acute for N00N-state metrology, where N00N states constitute the paradigmatic resource for super-resolution and Heisenberg scaling in lossless interferometry Lee2002Rosetta ; Mitchell2004 ; Nagata2007 ; Huver2008 ; Hiekkamaki2021 ; Winsten2024 , yet they are notoriously fragile to loss and decoherence Rubin2007 ; Demkowicz2009 ; Kolodynski2010 ; Datta2011 . Fundamental bounds have established that in the presence of noise the Heisenberg scaling cannot be maintained and that the achievable precision is ultimately limited to, at best, a constant-factor improvement over the SQL Dorner2009 ; Demkowicz2009 ; Escher2011 ; Datta2011 ; ThomasPeter2011 . This fragility has stimulated the search for more robust probes and control strategies holland_interferometric_1993 ; Huver2008 ; Thekkadath2020 ; Qin2023 ; Nielsen2023 , as well as the optimization of imperfect interferometers by deliberately exploiting asymmetries, for instance through unbalanced beam splitters or tailored input states, in order to maximize the (quantum) Fisher information in lossy and unbalanced settings Lee2009Loss ; AtamanMishra2024 ; HuLu2025 . A crucial point, often overlooked in experimental practice, is that high fringe visibility does not necessarily imply high metrological sensitivity Braunstein1994 ; Datta2011 ; Oh2017 . Visibility is governed by the coherence between the two interferometric paths and can, in many situations, be restored by compensating amplitude imbalance Lee2009Loss ; Agarwal2010 . The Fisher information, by contrast, depends not only on the slope of the interference signal, but also on the absolute detection probabilities, and is therefore sensitive to the conversion of relative losses into an effective global attenuation Demkowicz2009 ; Kolodynski2010 ; AtamanMishra2024 . As a result, the operating point that maximizes contrast does not, in general, coincide with the one that maximizes information. This distinction becomes even more relevant in a multiparameter perspective, where phase and loss (or phase and visibility) are not necessarily compatible parameters and where probe and measurement optimization for one quantity may be suboptimal for the other Bezerra2025 ; HuertaAlderete2022 . In this work we present an analytical, experimentally relevant study of these trade-offs for a Mach-Zehnder-like folded Franson interferometer fed with partially entangled N00N states, parametrized by a tunable input imbalance. We focus on two dominant and experimentally relevant imperfections: asymmetric losses between the interferometer arms and non-maximal path entanglement, equivalently an unbalanced input beam splitter Lee2009Loss ; AtamanMishra2024 . From exact detection probabilities we derive closed-form expressions for the N -photon fringe visibility and for the (normalized) Fisher information. We show that losses and input imbalance can compensate each other at the level of interference contrast: by appropriately tuning the entanglement, perfect visibility can be restored even in the presence of strong asymmetrical losses Agarwal2010 ; Lee2009Loss . However, this compensation does not, in general, maximize the Fisher information Demkowicz2009 ; AtamanMishra2024 . The operating point that optimizes contrast differs from that which optimizes sensitivity, reflecting the fact that restoring coherence does not compensate for the information loss associated with reduced detection rates. By benchmarking against single-photon strategies under identical loss conditions ThomasPeter2011 ; Oh2017 , we identify the parameter regimes in which multi-photon N00N interference still provides a genuine quantum advantage Qin2023 ; Nielsen2023 ; Thekkadath2020 ; Wang2024 , and we delineate which experimental parameters act as critical limitations and which can be used as soft control knobs for performance optimization. II Model The optical configuration considered in this work is shown in Fig. 1 and corresponds to a folded Franson interferometer Franson1989 ; Rarity1990 . The input state is generated in the ”State Preparation” stage and consists of a partially entangled path-encoded N -photon state of the form |\psi^{\mathrm{in}}_{N}\rangle_{ab}=\Bigl[\sqrt{\alpha}\,(\hat{a}^{\dagger})^{N}+\sqrt{1-\alpha}\,(\hat{b}^{\dagger})^{N}\Bigr]|00\rangle_{ab}, (1) where N denotes the total photon number and the modes a and b label the lower and upper arms of the interferometer, respectively. The parameter \alpha\in[0,1] controls the degree of path entanglement, with \alpha=1/2 corresponding to a maximally entangled N00N state. Throughout this work we assume the input state to be pure i.e. \text{Tr}(\rho^{2})=1 where \rho is the density matrix of the input state. Note that for N=1 the state reduces to a nonclassical superposition of the two interferometer paths, which we use as a reference case throughout this work. Figure 1: General scheme of the interferometric configuration considered in this work, based on a N00N state propagating in a Mach–Zehnder–like folded Franson interferometer. The two arrows represent the lower and upper arms of the interferometer. A relative phase shift and relative losses are introduced in the upper arm and are modeled by the unitary transformations \hat{U}_{\phi} and \hat{U}_{L} , respectively. Figure 2: (a) Fringe visibility [Eq. ( 8 )] as a function of the entanglement parameter \alpha for fixed relative losses ( p=0.3 ), for N=1,2, and 5 photons (blue, red, and green curves, respectively). Dotted lines indicate the global maxima, showing that the visibility can