⚛️ quantum physics

Teleportation-based squeezer for bosonic cluster states

This paper proposes and evaluates a novel teleportation-based squeezer for bosonic cluster states that utilizes unbalanced beam splitters and homodyne detection with unity-gain feed-forward, demonstrating superior performance over existing methods for both Gaussian and non-Gaussian input states to enable low-noise, scalable one-way quantum computation.

Original authors: Michal Matulík, Radim Filip, Petr Marek

Published 2026-07-07

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Michal Matulík, Radim Filip, Petr Marek

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

The Big Picture: Building a Quantum Computer with Light

Imagine you are trying to build a super-computer that uses light (photons) instead of electricity. This paper is about a specific tool needed to make that computer work: a "squeezer."

In the world of quantum light, information is carried in waves. Sometimes, to process information, you need to "squeeze" these waves. Think of a wave like a balloon. Squeezing it means making it very thin in one direction (so it's very precise) but making it very fat in the other direction (so it's very uncertain). This trade-off is a fundamental rule of quantum physics.

The authors are trying to figure out the best way to build this "squeezer" machine inside a quantum computer that works by teleporting information from one place to another.

The Problem: Noise and Messy Tools

In a perfect world, you could squeeze a light wave perfectly. But in the real world, everything is messy.

The "Static": Just like an old radio has static, quantum light has "noise" (random jitters).
The "Leaky Pipes": When light travels through glass or mirrors, some of it gets lost or gets mixed with the wrong stuff.

If your squeezer adds too much noise, the delicate quantum information gets ruined, and the computer stops working. The goal of this paper is to find the squeezer design that adds the least amount of noise.

The Two Competing Designs

The researchers compared two different ways to build this squeezer using a "teleportation" setup (where information is moved from one beam of light to another).

1. The "Phase-Shift" Squeezer (PS-sq)

How it works: Imagine you have a balanced seesaw (a beam splitter that splits light 50/50). To squeeze the light, you twist the angle of the light waves slightly before they hit the seesaw. It's like turning a dial to change the phase.
The Paper's Finding: This method is simple to build (it just needs simple mirrors), but it's not very good at squeezing without adding extra noise. It's like trying to cut a diamond with a dull knife; you can do it, but the result is rough.

2. The "Beam Splitter" Squeezer (BS-sq)

How it works: Instead of just twisting the angle, this method uses a special, unbalanced seesaw. Imagine a seesaw that splits the light 70/30 instead of 50/50. By carefully adjusting how much light goes to each side, you can achieve the squeeze.
The Paper's Finding: This method is more complex to build because you need to tune the "split" ratio perfectly. However, the paper proves that this method is much better. It squeezes the light more cleanly and adds significantly less noise than the first method.

The "Hybrid" Option

The researchers also looked at a third option that combines both twisting the angle and using an unbalanced seesaw. They found that this "super-combo" method performs exactly as well as the unbalanced seesaw method alone. So, you don't need the extra complexity of twisting the angle if you just use the unbalanced seesaw correctly.

The Test Drive: Vacuum vs. Single Photon

To see which squeezer was better, they tested them on two types of "cargo":

The Vacuum (Empty Space): This is like sending an empty box. It's a simple, smooth wave.
The Single Photon: This is like sending a single, fragile marble. This is a "non-Gaussian" state, meaning it has a weird, jagged shape that is very hard to preserve.

The Results:

For the Empty Box: The unbalanced seesaw (BS-sq) was clearly the winner, keeping the box much cleaner.
For the Fragile Marble: This was the real test. The "Phase-Shift" method (PS-sq) was so noisy that it often destroyed the special "jagged" shape of the marble, turning it into a boring, smooth blob. The unbalanced seesaw (BS-sq) was much better at keeping the marble's unique shape intact, especially when the equipment wasn't perfect (which is always the case in real life).

The Bottom Line

The paper concludes that if you want to build a quantum computer using light, you should stop using the simple "twist the angle" method. Instead, you should use the "unbalanced beam splitter" method.

Even though the unbalanced method requires more precise tuning of the equipment, it is the superior choice because it preserves the delicate quantum information much better. It's the difference between using a dull knife and a sharp one: the sharp one takes a bit more skill to handle, but it cuts the diamond perfectly without shattering it.

In short: The paper found a better recipe for a quantum "squeezer" that makes less noise and keeps quantum information safer, paving the way for more reliable quantum computers.

Title: Teleportation-based squeezer for bosonic cluster states
Authors: Michal Matulík, Radim Filip, and Petr Marek

Problem Statement

One-way quantum computation utilizing bosonic modes of light offers unparalleled scalability but requires robust, low-error Gaussian operations, specifically the squeezing gate. While phase shifts, displacements, and beam splitters are passive and easily integrated into cluster states, squeezing is an active operation. In measurement-based architectures, squeezing is typically implemented via teleportation protocols. However, existing methods often rely on fixed balanced beam splitters and optimally rotated homodyne detection, which may not yield optimal performance under realistic experimental conditions involving finite squeezing resources, loss, and detector inefficiencies. The paper addresses the need to identify and optimize the most effective teleportation-based squeezing gate for cluster state architectures, specifically evaluating how well these gates preserve the non-Gaussian nature of input states, which is crucial for universal quantum computing.

Methodology

The authors analyze and compare three distinct teleportation-based squeezing protocols using the Heisenberg formalism to derive noise characteristics and state transformations. The general setup involves a two-mode squeezed vacuum (TMSV) resource created from orthogonally squeezed vacuum states, mixed with an input state via beam splitters, followed by homodyne detection and unity-gain feed-forward displacement.

The three protocols compared are:

Phase-Shift Squeezer (PS-sq): Uses balanced beam splitters ( $t_1 = t_2 = 1/\sqrt{2}$ ) but introduces non-zero phase shifts ( $\phi, \zeta, \epsilon$ ) to implement shear and squeezing.
Beam Splitter Squeezer (BS-sq): Keeps phase shifts at zero but utilizes unbalanced beam splitters with variable transmission coefficients ( $t_1, t_2$ ) to achieve squeezing.
Combined Squeezer (BSPS-sq): A general scheme combining both variable beam splitters and phase shifts.

The study evaluates these protocols under two scenarios:

Ideal (Lossless): Pure resource states and unit-efficiency detectors.
Realistic: Finite squeezing resources (3, 6, and 9 dB), resource loss ( $\eta_S = 0.8$ ), and limited detector efficiency ( $\eta_H = 0.9$ ).

The performance is benchmarked using two input states:

Vacuum state: To test basic noise properties.
Single photon state: To assess the preservation of non-Gaussian features (Wigner negativity).

Key metrics include:

Fidelity: Between the transformed state and the ideally squeezed state.
Wigner Function Minimum ( $W(0,0)$ ): To quantify non-Gaussianity for the single photon state.
Total Added Noise ( $N_T$ ): Calculated after re-parameterizing for the squeezing factor.
Entanglement Breaking Bound ( $N_P$ ): The product of noise variances.
Genuine Non-Gaussianity (GnG): To measure the non-Gaussian nature based on photo-statistics.

Key Contributions

Optimization of Squeezing Gates: The paper identifies that the BS-sq protocol (using unbalanced beam splitters) is superior to the traditional PS-sq protocol (using phase shifts) across various resource levels and experimental imperfections.
Unified Framework: The authors provide a comprehensive analytical framework (Eqs. 1–13) describing the noise matrices and output states for all three protocols, including the effects of finite squeezing, loss, and detector inefficiency.
Performance Benchmarking: The study demonstrates that while the BSPS-sq (combined) scheme is the most general, the BS-sq scheme matches its performance, suggesting that variable beam splitters alone are sufficient for optimal squeezing without the need for complex phase adjustments.
Non-Gaussian Preservation Analysis: The work explicitly evaluates how different squeezing implementations degrade the non-Gaussian nature of single-photon states, a critical factor for quantum computational advantage.

Results

Fidelity: The BS-sq protocol consistently outperforms PS-sq in both ideal and realistic scenarios. In a realistic scenario with 9 dB resource squeezing, BS-sq achieves approximately a 16% higher fidelity for vacuum states at high target squeezing ( $10 \log s^2 \approx -9$ dB). For single photon states, the fidelity gain is up to 8% for 9 dB resource squeezing and 2% for 3 dB.
Wigner Negativity: BS-sq better preserves the negativity of the Wigner function for single photon states compared to PS-sq. However, the authors note that if resource squeezing is low (e.g., 3 dB in lossless or 6 dB in realistic scenarios), the negativity is lost regardless of the protocol.
Noise Characteristics: While the uncorrected total noise appears lower for PS-sq, this is an artifact of the squeezing parameter definition. After correcting for the squeezing factor, the BS-sq protocol exhibits lower total added noise.
Genuine Non-Gaussianity (GnG): Interestingly, while BS-sq offers higher fidelity, it leads to a quicker loss of non-Gaussianity as certified by GnG. This suggests that BS-sq states contain higher vacuum contributions in their photo-statistics compared to PS-sq, though BS-sq remains superior in terms of fidelity and Wigner negativity.
BAS-sq Performance: The "Basic Squeezing" (BAS-sq) approach, which mixes the input with a third squeezed state, introduces significantly more noise and performs poorly, confirming that the teleportation-based approach is superior.

Significance

The paper claims that the BS-sq approach opens a road to low-noise squeezing gates in experimentally achievable cluster states. By utilizing unbalanced beam splitters rather than relying solely on phase shifts, experimentalists can achieve better fidelity and better preservation of non-Gaussian features. The authors emphasize that while the PS-sq protocol is simpler to implement (requiring only phase shifters), this simplicity comes at the cost of performance. The results suggest that for near-term experimental tests, particularly where dynamic control of squeezing is not required, the BS-sq protocol is the preferred method. The study also highlights that the advantage of BS-sq is most pronounced in regimes of high target squeezing and that the performance gap narrows as resource squeezing increases, eventually merging in the limit of infinite resources.