⚛️ quantum physics

Precision Limits and Non-Gaussian Advantage in Optical Phase Estimation under Preparation Constraints

This paper establishes that by introducing a physically motivated envelope constraint on photon-number distributions to define a fixed resource quantity ( $N_\mathrm{RMS}$ ), non-Gaussian optical probes can achieve a genuine metrological advantage over Gaussian probes by attaining a finite Heisenberg-limited bound on the quantum Fisher information even at fixed mean photon number.

Original author: Víctor Izquierdo, Javier Lalueza-Puértolas

Published 2026-11-05

📖 4 min read🧠 Deep dive

CC BY 4.0

Original author: Víctor Izquierdo, Javier Lalueza-Puértolas

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the author. For technical accuracy, refer to the original paper. Read full disclaimer

In the world of light, measuring things with extreme precision is a game of counting photons, the tiny packets that make up a beam. For decades, scientists have used the average number of these packets as the standard ruler for how much effort a measurement requires. If you want to measure a tiny shift in the phase of a light wave—like detecting a gravitational ripple or a subtle change in a material—you shine a probe beam at it and see how the light changes. The more photons you use on average, the better your measurement becomes, but there is a ceiling. Classical light, like that from a standard laser, hits a limit known as the shot-noise barrier, where precision improves only in proportion to the square root of the number of photons. To break past this, researchers turned to quantum mechanics, using special "squeezed" states of light that can improve precision much faster, a feat known as the Heisenberg limit. However, a confusing problem emerged in the math: if you look only at the average number of photons, certain exotic, non-classical light states appear to offer infinite precision, suggesting you could measure perfectly without using any extra energy. This seemed to break the laws of physics, implying that the average photon count was a flawed way to measure the true cost of these experiments.

A team of researchers at the Autonomous University of Barcelona has resolved this paradox by looking at how these special states of light are actually made. They realized that the mathematical trick allowing for infinite precision relied on creating states with extremely rare, high-energy components that no real-world machine could ever produce. In the real world, every light source has physical limits; a laser cannot pump out infinite energy, and a device cannot generate a photon distribution that decays too slowly. The researchers introduced a new way to measure resources that accounts for these physical boundaries. Instead of just counting the average number of photons, they defined a "photon budget" based on the envelope, or the outer limit, of how likely the light is to contain a certain number of photons. This envelope is determined by the capabilities of the equipment used to create the light. By applying this realistic constraint, they proved that the mysterious infinite precision disappears. When you respect the physical limits of the machine, the precision is finite and follows a clear, predictable rule.

The study reveals a surprising advantage for using complex, non-standard light states over the familiar, smooth waves of classical or "Gaussian" light. When the researchers applied their new budget rule, they found that standard Gaussian probes, even when optimized, could only improve their precision linearly as the resource budget increased. In contrast, the more complex, non-Gaussian probes could achieve a quadratic improvement, meaning they get significantly more precise for the same amount of effort. This is a crucial distinction because it shows that under realistic experimental conditions, the exotic states are not just theoretically interesting but actually necessary to reach the highest possible levels of measurement accuracy. The researchers demonstrated this using a model of a universal set of optical tools, showing that the limits of these tools naturally create the envelope that separates the achievable from the impossible.

This work does more than just fix a mathematical inconsistency; it provides a practical guide for building better sensors. By understanding that the true cost of a measurement is defined by the physical limits of the preparation equipment, scientists can now design experiments that truly maximize their potential. The findings confirm that while the dream of infinite precision is impossible, the path to the best possible precision requires moving beyond simple light sources. The researchers showed that by carefully managing the photon budget within the constraints of real hardware, non-Gaussian states can outperform their classical counterparts by a significant margin. This establishes a new benchmark for quantum metrology, ensuring that future experiments are judged not by idealized numbers, but by the tangible resources available in the laboratory. The result is a clearer, more grounded understanding of what is possible when we push light to its limits.

Technical Summary: Precision Limits and Non-Gaussian Advantage in Optical Phase Estimation under Preparation Constraints

Problem Statement
In optical quantum metrology, the Quantum Fisher Information (QFI), $F_\theta$ , serves as the standard benchmark for estimation precision, bounded by the quantum Cramér–Rao bound. While Gaussian probes (e.g., squeezed-vacuum states) achieve Heisenberg scaling ( $F_\theta \propto \langle \hat{n} \rangle^2$ ) with the mean photon number $\langle \hat{n} \rangle$ , it is theoretically established that non-Gaussian probes can exhibit arbitrarily large QFI even at a fixed mean photon number. This creates a paradox where the mean photon number, traditionally used as the sole resource metric, fails to provide a finite universal bound on sensitivity. The authors argue that this measure obscures the resources associated with rare, highly excited components of the state, leading to an incomplete accounting of metrological resources. The central problem is to identify the physical constraints that render such divergent states inaccessible and to establish a rigorous resource framework that yields finite precision limits.

Methodology
The authors propose a framework based on preparation constraints derived from the physical limitations of state-generation methods. Instead of relying solely on the mean photon number, they introduce a photon-number distribution envelope, $(g_n)_n$ , which acts as an upper bound on the admissible photon-number probabilities $q_n = |\langle n|\psi\rangle|^2$ .

Envelope Definition: The envelope is defined such that $g_n = 1$ for $1 \le n \le N_{\text{max}}$ (a "free window") and $0 \le g_n \le 1$ for $n > N_{\text{max}}$ , with the condition that the second moment of the tail converges ( $\sum_{n \ge N_{\text{max}}} n^2 g_n < \infty$ ). This models physical realities such as finite pumping or bounded squeezing, where the probability of populating high-energy Fock states cannot decay slower than the source's own distribution.
Resource Quantification: Based on this envelope, the authors define a new resource quantity: the Root-Mean-Square (RMS) Photon Budget, $N_{\text{RMS}}(g)$ . This is defined as the square root of the maximum possible second moment allowed by the envelope:
$N_{\text{RMS}}(g) := \left( \max_{q \in Q_g} E_q[\hat{n}^2] \right)^{1/2}$
where $Q_g$ is the set of all admissible distributions bounded by $g$ .
Comparative Analysis: The authors derive upper bounds on the QFI for both general admissible probes and specifically for Gaussian probes subject to the same envelope. They analyze families of envelopes with algebraic and exponential decays to compare the scaling of $F_\theta$ against $N_{\text{RMS}}$ .
Physical Model: To ground the framework, they apply it to a preparation model using a universal optical gate set (displacements, squeezing, rotations, and Kerr gates), quantifying resources by total squeezing and displacement to determine the resulting envelope.

Key Contributions and Results

Finite Heisenberg Limit: The authors derive a finite upper bound on the QFI for any admissible probe: $F_\theta \le 4 N_{\text{RMS}}^2(g)$ . This recovers Heisenberg scaling in terms of the RMS budget, resolving the issue of divergent QFI at fixed mean photon number. The divergence previously observed in "heavy-tailed" or "asymmetrically superposed" states is shown to require an unbounded increase in $N_{\text{RMS}}$ , revealing a hidden resource cost.
Non-Gaussian Advantage Criteria: The paper establishes sufficient conditions under which non-Gaussian probes outperform all admissible Gaussian probes under the same preparation constraints:
- Algebraic Tails: For envelopes with algebraic decay $g_n \propto n^{-p}$ ( $p>3$ ), the ratio of non-Gaussian to Gaussian QFI exceeds unity by a constant factor only for a specific narrow range of exponents ( $3 < p < 2 + \sqrt{8/7}$ ) and sufficiently large $N_{\text{max}}$ . Crucially, this advantage manifests as a constant-factor improvement rather than a scaling separation in $N_{\text{RMS}}(g)$ .
- Exponential Tails: For envelopes with exponential decay (characterized by a tail length $\ell_g$ ), the authors prove a fundamental separation. The QFI of Gaussian probes is bounded by $F_\theta(\psi_G) \le 2\ell_g^2 + 4(1+2\ell_g)N_{\text{RMS}}$ , which grows at most linearly with $N_{\text{RMS}}$ . In contrast, optimal non-Gaussian probes achieve quadratic scaling ( $F_\theta \propto N_{\text{RMS}}^2$ ).
Mechanism of Advantage: The advantage arises because the envelope constraint limits the achievable squeezing parameter for Gaussian states (via the decay rate of the tail) while still permitting large second moments through non-Gaussian superpositions. Thus, Gaussian probes are restricted to shot-noise-like scaling in $N_{\text{RMS}}$ under fixed exponential decay, whereas non-Gaussian probes can attain the Heisenberg limit.
Universal Gate Model Application: In the context of a universal gate set, the authors demonstrate that experimental resources (total displacement and squeezing) directly determine the envelope parameters ( $N_{\text{max}}$ and decay rate). They show that non-Gaussian states (e.g., displaced cat states or superpositions of vacuum and high-Fock states) can achieve $F_\theta \sim N_{\text{RMS}}^2$ , while squeezed vacuum states suffer from a prefactor that decreases with increased squeezing and tolerance requirements.

Significance and Claims
The paper claims to resolve the ambiguity surrounding the use of mean photon number as a resource metric by introducing a physically motivated constraint that separates admissible states from those that are physically inaccessible. By defining the RMS photon budget, the authors provide a finite, attainable Heisenberg limit that accounts for the energy variance of the probe.

The primary significance lies in demonstrating that preparation constraints can make non-Gaussianity necessary to achieve Heisenberg scaling. Specifically, under constraints with fixed exponential tail decay, Gaussian probes are fundamentally limited to linear scaling with the resource budget, while non-Gaussian probes can achieve quadratic scaling. The work suggests that the "non-Gaussian advantage" is not merely a theoretical curiosity but a requirement imposed by the physical limitations of state generation when aiming for ultimate precision. The authors conclude that their framework connects abstract metrological bounds to the capabilities of actual optical setups, providing benchmarks based on available preparation resources.

Technical Summary: Precision Limits and Non-Gaussian Advantage in Optical Phase Estimation under Preparation Constraints

More like this