⚛️ quantum physics

Full-Period Optical Phase Estimation with Heisenberg Scaling Using Displaced Squeezed States and Gaussian Measurements

This paper proposes a two-stage Gaussian strategy for full-period optical phase estimation that utilizes displaced squeezed states and heterodyne measurements for coarse localization followed by adaptive homodyne measurements on squeezed-vacuum probes, achieving Heisenberg scaling within a fixed energy constraint while significantly outperforming coherent-state-based approaches.

Original authors: Marco A. Rodríguez-García, Luis Medina-Dozal, Francisco E. Becerra

Published 2026-07-07

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Marco A. Rodríguez-García, Luis Medina-Dozal, Francisco E. Becerra

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 authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to find a specific, hidden treasure on a giant, circular island. The island represents all possible angles (from 0 to 360 degrees). Your goal is to pinpoint the exact location of the treasure using a limited amount of "energy" (like a battery charge or a budget of light particles called photons).

This paper proposes a clever two-step strategy to find that treasure as accurately as possible, beating the limits of what standard tools can do.

The Problem: The "Blind" Search

Usually, if you try to guess a location on a circle using standard light (coherent states), your guess gets better as you use more energy, but only at a slow, steady pace. It's like walking around the island with a dim flashlight; you get closer, but it takes a long time to be sure.

However, there is a special type of light called squeezed light. Think of squeezed light as a super-sharp, laser-focused beam that can pinpoint a location with incredible precision—much faster than the standard method. But there's a catch: squeezed light has a "blind spot." Because of its unique shape, it looks exactly the same if you rotate it halfway around the circle (180 degrees). It can't tell the difference between "North" and "South." If you use only squeezed light, you might find the treasure, but you won't know if it's actually on the opposite side of the island.

The Solution: A Two-Stage Detective Strategy

The authors propose a "Two-Stage" plan to solve this. They split their total energy budget between two different detectives working together.

Stage 1: The Rough Sketch (Coarse Localization)

The Tool: A slightly "squashed" and "pushed" light beam (Displaced Squeezed State).
The Job: To figure out which half of the island the treasure is in.
The Analogy: Imagine you have a flashlight that isn't perfectly sharp, but it's bright enough to see the general neighborhood. You shine it around the island. Because this light is "displaced" (pushed off-center), it breaks the symmetry. It can tell the difference between North and South.

What happens: You use a small part of your energy to get a "rough sketch." You narrow the search down from the whole circle to a specific 90-degree slice (a quarter of the island).
The Risk: If you don't use enough energy here, you might guess the wrong slice. If you pick the wrong slice, your final answer will be wildly off, no matter how good the next step is. This is called an "overshoot penalty."

Stage 2: The Laser Focus (Local Estimation)

The Tool: Pure, high-powered squeezed light (Squeezed Vacuum).
The Job: To find the exact spot within that 90-degree slice.
The Analogy: Now that you know the treasure is in this specific quarter of the island, you switch to your super-sharp, high-tech laser. Because you already know the general area, the laser's "blind spot" (the North/South confusion) doesn't matter anymore. You can zoom in and find the exact coordinates with extreme precision.

The Result: This step uses the remaining energy to get a result that is incredibly accurate, far better than standard methods could ever achieve on the whole island.

The Big Discovery: How to Split the Budget

The paper does the math to figure out the perfect way to split your energy between the "Rough Sketch" (Stage 1) and the "Laser Focus" (Stage 2).

The Trade-off: If you spend too much energy on the rough sketch, you don't have enough left for the laser, and your final precision suffers. If you spend too little on the sketch, you might pick the wrong slice, and the laser's precision becomes useless.
The Sweet Spot: The authors found that using a little bit of the special "displaced" light in Stage 1 is actually better than using standard light. It helps you pick the right slice more reliably without costing too much energy.
The Outcome: Even with a modest amount of energy (like 25 photons) and realistic limits on how "sharp" the light can be (12 dB of squeezing), this two-step method gets results that are very close to the theoretical best possible limit (the "Heisenberg limit").

The Bottom Line

Think of this like finding a needle in a haystack.

Old Way: You use a standard magnet. It takes a long time to find the needle, and you might miss it if the haystack is huge.
The Paper's Way: First, you use a slightly weaker magnet to quickly find the bale of hay the needle is in (Stage 1). Then, you use a super-powerful, specialized magnet to find the needle inside that specific bale (Stage 2).

By combining a "good enough" rough guess with a "perfect" fine-tune, the authors show you can find the treasure on the whole circle with much higher precision than ever before, using the same amount of energy. They proved mathematically that this strategy works and calculated exactly how to balance the energy to get the best results.

Problem Statement

Optical quantum metrology aims to estimate physical parameters, specifically optical phase, with precision surpassing the Standard Quantum Limit (SQL). While squeezed vacuum states can achieve Heisenberg scaling ( $\sim 1/\bar{n}^2$ ) within identifiable phase intervals, they possess an inherent $\pi$ -rotation symmetry ( $\rho_{\theta+\pi} = \rho_\theta$ ). This symmetry prevents global phase estimation over the full circle $\phi \in [0, 2\pi)$ without prior information, as measurements cannot distinguish between $\theta$ and $\theta + \pi$ .

Existing approaches to full-period estimation either rely on complex, adaptive optimization of Gaussian probes at every step (which is experimentally daunting) or are limited to local intervals. The challenge is to design a practical protocol that achieves full-period identifiability while maintaining high precision (approaching Heisenberg scaling) under fixed energy constraints and experimental limitations on squeezing.

Methodology

The authors propose a two-stage optimized Gaussian protocol that separates the tasks of global localization and local high-precision estimation. The protocol operates under a fixed total energy constraint $E = E_1 + E_2$ (measured in mean photon numbers).

Stage I: Coarse Global Localization

Probe: Displaced squeezed states $|\alpha_1, \zeta_1\rangle$ with non-zero displacement ( $|\alpha_1| > 0$ ).
Measurement: Heterodyne detection.
Function: The non-zero displacement breaks the $\pi$ -symmetry of squeezed vacuum, restoring full-period identifiability. The heterodyne measurement provides a coarse estimate $\hat{\theta}_1$ and defines a localization window $W(Z)$ of length $\pi/2$ centered at $\hat{\theta}_1$ .
Goal: To select an interval containing the true phase with high probability, thereby enabling the use of locally optimal strategies in Stage II.

Stage II: Local Adaptive Estimation

Probe: Squeezed-vacuum states $|0, \zeta_2\rangle$ (no displacement).
Measurement: Adaptive homodyne detection, where the local oscillator phase is updated based on previous outcomes.
Function: Once the phase is localized within the $\pi/2$ window by Stage I, Stage II performs efficient estimation. The adaptive strategy keeps the measurement near the optimal operating point to approach the Quantum Cramér-Rao Bound (QCRB) for squeezed vacuum within that interval.

Theoretical Framework
The authors derive a Generalized Cramér-Rao Bound (CRB) for this two-stage class of strategies. Unlike standard local bounds, this generalized bound accounts for two distinct error contributions:

Local Estimation Error: The variance of the estimator in Stage II, conditioned on the event that Stage I successfully selected a window containing the true phase.
Overshoot Penalty: A geometric penalty term arising when Stage I fails to select the correct window (i.e., the true phase lies outside $W(Z)$ ). In this case, the final estimator is constrained to an incorrect arc, incurring a minimum error determined by the distance to the selected window. This term is independent of the Fisher information in Stage II.

Key Contributions

Generalized Bound Derivation: The paper establishes a rigorous lower bound (Theorem III.2) that explicitly separates the local estimation error from the "overshoot" penalty caused by localization failures. This provides a tractable framework for analyzing full-period estimation with finite resources.
Analysis of Stage I Probes:
- Coherent States: The authors derive the exact Rician phase density for the error distribution when using coherent states in Stage I, allowing for the calculation of coverage probability and overshoot penalties.
- Displaced Squeezed States: They analyze the performance of displaced squeezed probes. While these states offer higher local Fisher information (due to squeezing), the authors show that for a single probe, the energy cost of squeezing often outweighs the localization benefit compared to coherent states. However, for multiple probes ( $N_1 > 1$ ), displaced squeezed states can reduce the number of probes required to achieve a target coverage probability, though not necessarily the total energy cost in isolation.
Joint Optimization: The paper presents a numerical optimization of the two-stage protocol under a fixed total energy budget ( $E \le 25$ photons) and a maximum squeezing limit (12 dB). It optimizes the allocation of energy between Stage I (localization) and Stage II (estimation), as well as the number of probes ( $N_1, N_2$ ) and squeezing parameters.

Results

Performance vs. Ideal Limit: The optimized two-stage protocols using displaced squeezed states in Stage I consistently outperform protocols using coherent states in Stage I across the investigated energy range.
Proximity to Heisenberg Scaling: For total energies up to 25 photons and squeezing up to 12 dB, the optimized protocols achieve a mean squared error within a factor of 3–30 of the idealized local QCRB for squeezed vacuum (which assumes the phase is already known to be within a small interval).
Energy Allocation: The optimization reveals that the best performance is achieved by allocating a modest amount of squeezing to Stage I (primarily to break symmetry and improve localization) while concentrating the majority of the squeezing resource in Stage II to maximize local precision.
Role of $N_2$ : The number of probes in Stage II ( $N_2$ ) dictates the trade-off between squeezing per probe and the number of adaptive steps. A smaller $N_2$ (e.g., $N_2=1$ ) concentrates energy into a highly squeezed probe, approaching the local QCRB closely but requiring high experimental squeezing. Larger $N_2$ reduces the per-probe squeezing requirement (making it more experimentally feasible) but moves the performance further from the ideal local limit due to the dominance of linear (shot-noise) terms in the Fisher information at lower energies per probe.

Significance and Claims

The paper claims to provide a practical solution for full-period optical phase estimation that bridges the gap between theoretical Heisenberg scaling and experimental feasibility.

Practicality: By restricting adaptivity to Gaussian measurements (heterodyne and homodyne) and using fixed families of probe states, the protocol avoids the "daunting" complexity of optimizing probe parameters at every step, making it suitable for current experimental platforms.
Robustness: The derived generalized bound explicitly handles the risk of localization failure (overshoot), a critical factor often ignored in purely local estimation theories.
Resource Efficiency: The study demonstrates that even with moderate resources (low photon numbers and limited squeezing), displaced squeezed states in the localization stage can significantly improve the overall estimation bound compared to classical coherent states, bringing the system closer to the fundamental quantum limits of phase estimation.

The authors conclude that while the protocol does not reach the idealized local QCRB (due to the necessary cost of global localization), it offers a viable path to sub-SQL precision over the full phase circle, with performance scaling favorably as total energy increases.