Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2610.02625

Gaussian Fisher Information Is Superadditive

Jiaxin Liu, Zuoxian Wang, Danyue Ma

quant-phSubmitted Oct 2, 2026• Announced Oct 5, 2026• 1 citation
arXivHTMLPDF

Abstract

Quantum Fisher information adds over independent probes, so a single parameter is best measured probe by probe. We show that Gaussian measurements, the linear optics and homodyne detection of optical and microwave experiments, break this rule: two independent modes are better measured together. The uncertainty principle leaves a linear detector half of phase space, and for two modes the choice of half is a resource. Bell homodyne, a balanced beam splitter followed by two homodyne detectors, beats the best separate readout when the modes differ in quadrature width and in how the width responds to the parameter. Heterodyne detection splits a mode on a beam splitter to read both quadratures and pays one unit of vacuum noise for the empty port. Bell homodyne fills that port with the second mode, so the noise turns into signal. We prove that the gain stays below $(\sqrt2-1)^2=17.157\%$ for every parameter carried by widths. For thermal modes at one temperature, the canonical case, a gain was conjectured impossible. We prove that any frequency difference opens a temperature window and that the gain peaks at $12.699\%$ at frequency ratio $3.318$, where Bell homodyne is the optimal Gaussian measurement. Standard hardware reaches the gain: two coupled resonators with two homodyne detectors, or a phase-preserving amplifier whose idler band is fed by the same thermal source.

AI Summary

Can collective Gaussian measurements improve single-parameter estimation from independent Gaussian probes, despite additivity of unrestricted quantum Fisher information? This paper studies two phase-insensitive modes whose parameter dependence resides in their quadrature widths, particularly unequal-frequency thermal modes at a common temperature. It gives a counterexample to a conjecture that joint Gaussian thermometry cannot outperform optimal separate Gaussian readout. The single-mode optimum is homodyne or heterodyne, switching at variance $\nu_c=1+\sqrt2$ in vacuum units. Balanced Bell homodyne instead measures two commuting joint quadratures and can exceed the sum of those separate optima. For thermal modes, every nonzero frequency splitting admits a temperature window with an advantage. The maximum Bell-over-local Fisher-information gain is $12.699322\%$, at frequency ratio $3.31798$, with the lower-frequency mode exactly at the homodyne–heterodyne crossover. For arbitrary width-response laws, the corresponding Bell gain has supremum $(\sqrt2-1)^2\approx17.157\%$. Analytic optimization establishes the local benchmark, small-splitting onset law and width-law bound. Interval-arithmetic Bernstein certificates establish Bell optimality over all Gaussian measurements on two parameter rectangles containing the thermal maximum. Proposed implementations use coupled resonators or thermally populated amplifier signal/idler bands; at the maximum, Bell beats equally inefficient separate homodyne/heterodyne readout above $57.95\%$ efficiency. The universal bound is proved for Bell readout, not arbitrary multimode Gaussian measurements. Passive-network pairing optimality is proved through three modes; larger-mode claims have numerical support only. General two-mode Gaussian optimality outside the certified regions remains unproved. The results are theoretical, and photon counting still reaches substantially higher Fisher information.

Some signs of AI-generated textOverclaims

Central claim may not hold: The stated universal 17.157% collective-gain cap is proved for Bell homodyne, but the paper does not establish it for arbitrary joint Gaussian measurements and arbitrary width responses. The stated 12.699% peak thermal collective gain is globally optimized only within Bell homodyne; Gaussian optimality on two thermal rectangles does not exclude a larger gain from another Gaussian receiver elsewhere.

Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.

Comments
32 pages, 10 figures

Related papers

Quantum sensing with a driven oscillator that returns to the vacuum

Jesús Fuentes Aguilar

A weak force displaces a quantum oscillator in phase space, and the quantum Fisher information sets how precisely that displacement can be measured. The usual way to raise it is to squeeze the probe, but squeezing is fragile, has to be remade each cycle, and fades as the oscillator loses energy. We propose instead a quadratic Hamiltonian, varying in time, whose propagator traces a closed loop in the symplectic group $$\textrm{Sp}(2,\mathbb {R})$$ , the probe passes through strong parametric gain and returns to the vacuum with no squeezing left. The gain enlarges the response to the signal instead of lowering the noise, and because the probe ends where it began, cycles follow one another with nothing to prepare in between, one quadrature being read at the end by ordinary homodyne detection. One smooth function carries the whole propagator, thus a loop is designed by choosing it and reading off the drive it implies, and the Fisher information is two integrals of the signal against the propagator. We give a family of such loops in closed form and three upper bounds, the sharpest reached exactly by the signal matched to the loop. When the oscillator also loses photons to a bath at rate $$\kappa $$ , the Fisher information $$\mathcal {F}$$ over a cycle from $$t_{0}$$ to T obeys $$\mathcal {F} \le 4\textrm{e}^{\kappa (T-t_{0})}\Vert \tilde{s}\Vert ^{2}[(T-t_{0}) + \eta /\kappa ]$$ , where $$\Vert \tilde{s}\Vert ^{2}$$ is the signal energy and $$\eta $$ the noise the bath injects while the gain is on. The inequality has the shape of a second law: the first term is what a loop that merely rotates already achieves, the second is the price of the gain.

Discover Quantum Science•Oct 26, 2026•0 citations

Practical resources and measurements for lossy optical quantum metrology

Changhun Oh, Su-Yong Lee, Hyunchul Nha +1 more

We study the sensitivity of phase estimation in a lossy Mach-Zehnder interferometer (MZI) using two general, and practical, resources generated by a laser and a nonlinear optical medium with passive optimal elements, which are readily available in the laboratory: One is a two-mode separable coherent and squeezed vacuum state at a beam splitter and the other is a two-mode squeezed vacuum state. In view of the ultimate precision given by quantum Fisher information, we show that the two-mode squeezed vacuum state can achieve a lower bound of estimation error than the coherent and squeezed vacuum state under a photon-loss channel. We further consider practical measurement schemes, homodyne detection and photon number resolving detection (PNRD), to characterize the accuracy of phase estimation in reality and find that the coherent and squeezed vacuum state largely achieves a lower bound than the two-mode squeezed vacuum in the lossy MZI while maintaining quantum enhancement over the shot-noise limit. By comparing homodyne detection and PNRD, we demonstrate that quadrature measurement with homodyne detection is more robust against photon loss than parity measurement with PNRD. We also show that double homodyne detection can provide a better tool for phase estimation than single homodyne detection against photon loss.

arXiv:1707.03210•Physical Review A•Dec 4, 2017•37 citations

Optimal initial states for quantum Fisher information in linearized cavity optomechanics

Wangjun Lu, Qing Yu, Ying Li +4 more

We find the optimal initial states of a linearized cavity-optomechanical system for estimating the single-photon coupling by quantum Fisher information. In the red-detuned (beam-splitter) regime the interaction exchanges excitations between the two modes, and for pure inputs the Fisher information is four times the variance of the excitation-exchange operator, reducing the optimization to variance maximization with a closed-form solution. With the mechanical mode in a Fock state the probe is degenerate: any state saturating the photon-number constraint is optimal, and a Fock reference with many phonons enhances the per-photon sensitivity linearly. For an arbitrary fixed mechanical reference it becomes a phase-matched squeezed vacuum, from an exact bound on that coherence obtained by a Lagrange-dual argument. Jointly optimizing both modes at fixed total excitation number, the quantum Fisher information obeys a Heisenberg bound quadratic in the energy, saturated by equal superpositions of the two extremal eigenstates of the generator, equivalently a two-mode NOON state after a balanced beam splitter; balanced product Fock states reach only a linear enhancement. Exact diagonalization in truncated Fock spaces confirms all the results, the familiar single-photon input being the simplest saturating case. Three extensions are worked out: the blue sideband has an unbounded two-mode-squeezing spectrum, so the entangled optimum becomes impossible and the balanced product Fock state is exactly optimal at definite energy; under cavity loss the information accumulation turns linear on the cavity-lifetime timescale, at a renewal-optimal duration of about two and a half lifetimes; and for a state-of-the-art quantum-coherent coupling experiment the per-hour relative precision lies between one part in a million and ten parts in a million, number and thermal references being the most robust.

quant-pharXiv:2610.01417•Oct 2, 2026•0 citations

Optimal Gaussian measurements for phase estimation in single-mode Gaussian metrology

Changhun Oh, Changhyoup Lee, Carsten Rockstuhl +4 more

Abstract The central issue in quantum parameter estimation is to find out the optimal measurement setup that leads to the ultimate lower bound of an estimation error. We address here a question of whether a Gaussian measurement scheme can achieve the ultimate bound for phase estimation in single-mode Gaussian metrology that exploits single-mode Gaussian probe states in a Gaussian environment. We identify three types of optimal Gaussian measurement setups yielding the maximal Fisher information depending on displacement, squeezing, and thermalization of the probe state. We show that the homodyne measurement attains the ultimate bound for both displaced thermal probe states and squeezed vacuum probe states, whereas for the other single-mode Gaussian probe states, the optimized Gaussian measurement cannot be the optimal setup, although they are sometimes nearly optimal. We then demonstrate that the measurement on the basis of the product quadrature operators $$\hat X\hat P + \hat P\hat X$$ X ̂ P ̂ + P ̂ X ̂ , i.e., a non-Gaussian measurement, is required to be fully optimal.

arXiv:1805.08495•npj Quantum Information•Jan 28, 2019•94 citations

Single-port homodyne detection achieves optimal multiphase sensing in an nnn-port ring cavity

Jiayu He, Aviral Kumar Pandey, Krishna Mohan Mishra +2 more

Estimating several phases simultaneously is more than estimating them one at a time: the optimal measurements for different parameters may not commute, and in a multiparameter problem the quantum Cramér-Rao bound is not automatically achievable. Ring cavities are a natural setting for this question, since a single coherent input can be made to interact repeatedly with every internal phase before it is read out. We study the simultaneous local estimation of the $n$ internal phases of a lossless $n$-port ring cavity probed by a fixed-energy multimode coherent state, optimizing over the full coherent-input space and over the operating point of the cavity. We identify a necessary and sufficient condition for the global optimum: the input energy should be distributed equally among the ports, and the internal phases and input phases satisfy a simple phase-matching condition for every port. The optimum is unique up to a common optical phase and minimizes both the symmetric logarithmic derivative (SLD) cost and the sloppiness of the model. At this point the quantum Fisher information matrix is full rank and its spectrum is obtained in closed form, the Uhlmann curvature vanishes, and no incompatibility penalty has to be paid. The precision retains the coherent-state scaling and the dependence on the number of phases, but the cavity feedback reduces the prefactor relative to $n$ independent single-pass experiments. Finally, we show that this bound may be achieved in practice: the Fisher information of a balanced homodyne measurement at each output port, with local-oscillator phases locked to the prepared input phases is equal to the optimal quantum Fisher information. Our results establish the ring cavity as a platform for optimal multiphase sensing and provide the ultimate benchmark against which nonclassical probes may be evaluated for further precision enhancement.

quant-pharXiv:2611.02754•Nov 3, 2026•0 citations

Noise in homodyne detection

Bonny L. Schumaker

A simple but rigorous analysis of the important sources of noise in homodyne detection is presented. Output noise and signal-to-noise ratios are compared for direct detection, conventional (one-port) homodyning, and two-port homodyning, in which one monitors both output ports of a 50-50 beam splitter. It is shown that two-port homodyning is insensitive to local-oscillator quadrature-phase noise and hence provides (1) a means of detecting reduced quadrature-phase fluctuations (squeezing) that is perhaps more practical than one-port homodyning and (2) an output signal-to-noise ratio that can be a modest to significant improvement over that of one-port homodyning and direct detection.

Optics Letters•May 1, 1984•244 citations
Quantum Fisher information adds over independent probes, so a single parameter is best measured probe by probe. We show that Gaussian measurements, the linear optics and homodyne detection of optical and microwave experiments, break this rule: two independent modes are better measured together. The uncertainty principle leaves a linear detector half of phase space, and for two modes the choice of half is a resource. Bell homodyne, a balanced beam splitter followed by two homodyne detectors, beats the best separate readout when the modes differ in quadrature width and in how the width responds to the parameter. Heterodyne detection splits a mode on a beam splitter to read both quadratures and pays one unit of vacuum noise for the empty port. Bell homodyne fills that port with the second mode, so the noise turns into signal. We prove that the gain stays below (2−1)2=17.157%(\sqrt2-1)^2=17.157\%(2​−1)2=17.157% for every parameter carried by widths. For thermal modes at one temperature, the canonical case, a gain was conjectured impossible. We prove that any frequency difference opens a temperature window and that the gain peaks at 12.699%12.699\%12.699% at frequency ratio 3.3183.3183.318, where Bell homodyne is the optimal Gaussian measurement. Standard hardware reaches the gain: two coupled resonators with two homodyne detectors, or a phase-preserving amplifier whose idler band is fed by the same thermal source.
Can collective Gaussian measurements improve single-parameter estimation from independent Gaussian probes, despite additivity of unrestricted quantum Fisher information? This paper studies two phase-insensitive modes whose parameter dependence resides in their quadrature widths, particularly unequal-frequency thermal modes at a common temperature. It gives a counterexample to a conjecture that joint Gaussian thermometry cannot outperform optimal separate Gaussian readout.
The single-mode optimum is homodyne or heterodyne, switching at variance νc=1+2\nu_c=1+\sqrt2νc​=1+2​ in vacuum units. Balanced Bell homodyne instead measures two commuting joint quadratures and can exceed the sum of those separate optima. For thermal modes, every nonzero frequency splitting admits a temperature window with an advantage. The maximum Bell-over-local Fisher-information gain is 12.699322%12.699322\%12.699322%, at frequency ratio 3.317983.317983.31798, with the lower-frequency mode exactly at the homodyne–heterodyne crossover. For arbitrary width-response laws, the corresponding Bell gain has supremum (2−1)2≈17.157%(\sqrt2-1)^2\approx17.157\%(2​−1)2≈17.157%.
Analytic optimization establishes the local benchmark, small-splitting onset law and width-law bound. Interval-arithmetic Bernstein certificates establish Bell optimality over all Gaussian measurements on two parameter rectangles containing the thermal maximum. Proposed implementations use coupled resonators or thermally populated amplifier signal/idler bands; at the maximum, Bell beats equally inefficient separate homodyne/heterodyne readout above 57.95%57.95\%57.95% efficiency.
The universal bound is proved for Bell readout, not arbitrary multimode Gaussian measurements. Passive-network pairing optimality is proved through three modes; larger-mode claims have numerical support only. General two-mode Gaussian optimality outside the certified regions remains unproved. The results are theoretical, and photon counting still reaches substantially higher Fisher information.
A weak force displaces a quantum oscillator in phase space, and the quantum Fisher information sets how precisely that displacement can be measured. The usual way to raise it is to squeeze the probe, but squeezing is fragile, has to be remade each cycle, and fades as the oscillator loses energy. We propose instead a quadratic Hamiltonian, varying in time, whose propagator traces a closed loop in the symplectic group Sp(2,R)\textrm{Sp}(2,\mathbb {R})Sp(2,R) , the probe passes through strong parametric gain and returns to the vacuum with no squeezing left. The gain enlarges the response to the signal instead of lowering the noise, and because the probe ends where it began, cycles follow one another with nothing to prepare in between, one quadrature being read at the end by ordinary homodyne detection. One smooth function carries the whole propagator, thus a loop is designed by choosing it and reading off the drive it implies, and the Fisher information is two integrals of the signal against the propagator. We give a family of such loops in closed form and three upper bounds, the sharpest reached exactly by the signal matched to the loop. When the oscillator also loses photons to a bath at rate κ\kappa κ , the Fisher information F\mathcal {F}F over a cycle from t0t_{0}t0​ to T obeys F≤4eκ(T−t0)∥s~∥2[(T−t0)+η/κ]\mathcal {F} \le 4\textrm{e}^{\kappa (T-t_{0})}\Vert \tilde{s}\Vert ^{2}[(T-t_{0}) + \eta /\kappa ]F≤4eκ(T−t0​)∥s~∥2[(T−t0​)+η/κ] , where ∥s~∥2\Vert \tilde{s}\Vert ^{2}∥s~∥2 is the signal energy and η\eta η the noise the bath injects while the gain is on. The inequality has the shape of a second law: the first term is what a loop that merely rotates already achieves, the second is the price of the gain.
We study the sensitivity of phase estimation in a lossy Mach-Zehnder interferometer (MZI) using two general, and practical, resources generated by a laser and a nonlinear optical medium with passive optimal elements, which are readily available in the laboratory: One is a two-mode separable coherent and squeezed vacuum state at a beam splitter and the other is a two-mode squeezed vacuum state. In view of the ultimate precision given by quantum Fisher information, we show that the two-mode squeezed vacuum state can achieve a lower bound of estimation error than the coherent and squeezed vacuum state under a photon-loss channel. We further consider practical measurement schemes, homodyne detection and photon number resolving detection (PNRD), to characterize the accuracy of phase estimation in reality and find that the coherent and squeezed vacuum state largely achieves a lower bound than the two-mode squeezed vacuum in the lossy MZI while maintaining quantum enhancement over the shot-noise limit. By comparing homodyne detection and PNRD, we demonstrate that quadrature measurement with homodyne detection is more robust against photon loss than parity measurement with PNRD. We also show that double homodyne detection can provide a better tool for phase estimation than single homodyne detection against photon loss.
We find the optimal initial states of a linearized cavity-optomechanical system for estimating the single-photon coupling by quantum Fisher information. In the red-detuned (beam-splitter) regime the interaction exchanges excitations between the two modes, and for pure inputs the Fisher information is four times the variance of the excitation-exchange operator, reducing the optimization to variance maximization with a closed-form solution. With the mechanical mode in a Fock state the probe is degenerate: any state saturating the photon-number constraint is optimal, and a Fock reference with many phonons enhances the per-photon sensitivity linearly. For an arbitrary fixed mechanical reference it becomes a phase-matched squeezed vacuum, from an exact bound on that coherence obtained by a Lagrange-dual argument. Jointly optimizing both modes at fixed total excitation number, the quantum Fisher information obeys a Heisenberg bound quadratic in the energy, saturated by equal superpositions of the two extremal eigenstates of the generator, equivalently a two-mode NOON state after a balanced beam splitter; balanced product Fock states reach only a linear enhancement. Exact diagonalization in truncated Fock spaces confirms all the results, the familiar single-photon input being the simplest saturating case. Three extensions are worked out: the blue sideband has an unbounded two-mode-squeezing spectrum, so the entangled optimum becomes impossible and the balanced product Fock state is exactly optimal at definite energy; under cavity loss the information accumulation turns linear on the cavity-lifetime timescale, at a renewal-optimal duration of about two and a half lifetimes; and for a state-of-the-art quantum-coherent coupling experiment the per-hour relative precision lies between one part in a million and ten parts in a million, number and thermal references being the most robust.
Abstract The central issue in quantum parameter estimation is to find out the optimal measurement setup that leads to the ultimate lower bound of an estimation error. We address here a question of whether a Gaussian measurement scheme can achieve the ultimate bound for phase estimation in single-mode Gaussian metrology that exploits single-mode Gaussian probe states in a Gaussian environment. We identify three types of optimal Gaussian measurement setups yielding the maximal Fisher information depending on displacement, squeezing, and thermalization of the probe state. We show that the homodyne measurement attains the ultimate bound for both displaced thermal probe states and squeezed vacuum probe states, whereas for the other single-mode Gaussian probe states, the optimized Gaussian measurement cannot be the optimal setup, although they are sometimes nearly optimal. We then demonstrate that the measurement on the basis of the product quadrature operators X^P^+P^X^\hat X\hat P + \hat P\hat XX^P^+P^X^ X ̂ P ̂ + P ̂ X ̂ , i.e., a non-Gaussian measurement, is required to be fully optimal.
Estimating several phases simultaneously is more than estimating them one at a time: the optimal measurements for different parameters may not commute, and in a multiparameter problem the quantum Cramér-Rao bound is not automatically achievable. Ring cavities are a natural setting for this question, since a single coherent input can be made to interact repeatedly with every internal phase before it is read out. We study the simultaneous local estimation of the nnn internal phases of a lossless nnn-port ring cavity probed by a fixed-energy multimode coherent state, optimizing over the full coherent-input space and over the operating point of the cavity. We identify a necessary and sufficient condition for the global optimum: the input energy should be distributed equally among the ports, and the internal phases and input phases satisfy a simple phase-matching condition for every port. The optimum is unique up to a common optical phase and minimizes both the symmetric logarithmic derivative (SLD) cost and the sloppiness of the model. At this point the quantum Fisher information matrix is full rank and its spectrum is obtained in closed form, the Uhlmann curvature vanishes, and no incompatibility penalty has to be paid. The precision retains the coherent-state scaling and the dependence on the number of phases, but the cavity feedback reduces the prefactor relative to nnn independent single-pass experiments. Finally, we show that this bound may be achieved in practice: the Fisher information of a balanced homodyne measurement at each output port, with local-oscillator phases locked to the prepared input phases is equal to the optimal quantum Fisher information. Our results establish the ring cavity as a platform for optimal multiphase sensing and provide the ultimate benchmark against which nonclassical probes may be evaluated for further precision enhancement.
A simple but rigorous analysis of the important sources of noise in homodyne detection is presented. Output noise and signal-to-noise ratios are compared for direct detection, conventional (one-port) homodyning, and two-port homodyning, in which one monitors both output ports of a 50-50 beam splitter. It is shown that two-port homodyning is insensitive to local-oscillator quadrature-phase noise and hence provides (1) a means of detecting reduced quadrature-phase fluctuations (squeezing) that is perhaps more practical than one-port homodyning and (2) an output signal-to-noise ratio that can be a modest to significant improvement over that of one-port homodyning and direct detection.