Composable free-space continuous-variable quantum key distribution using discrete modulation

This paper presents a composable, finite-size secure continuous-variable quantum key distribution system using discrete modulation and polarization encoding over urban atmospheric channels, validated by a laboratory demonstration that calculates key rates against collective attacks without Gaussian assumptions.

The Gist

Imagine you want to send a top-secret message to a friend, but you're worried a spy (let's call her Eve) might be listening in. In the world of quantum physics, there's a special way to do this called Quantum Key Distribution (QKD). It uses the weird rules of quantum mechanics to guarantee that if Eve tries to listen, she leaves a trace, and you know to throw away the message.

This paper describes a new, robust way to build this secret-keeping system, specifically designed for sending messages through the air (like between buildings in a city) rather than through glass fibers.

Here is the breakdown of their work using simple analogies:

1. The Problem: Sending Messages Through a Stormy Window

Most quantum systems today use glass fibers (like internet cables) to send light. But what if you want to send a secret message across a city square where there are no cables? You have to send it through the air.

• The Challenge: The air is turbulent. It's like looking through a wavy window or a heat haze. The wind and heat make the light jitter and twist.
• The Old Way: Previous systems tried to encode information in the brightness or phase of the light. In a stormy atmosphere, these get scrambled easily.
• The New Solution: This team used Polarization. Imagine light as a rope. You can shake it up-and-down or side-to-side. Even if the air twists the rope, it doesn't easily change the direction of the shake (unlike brightness). They used this "shake direction" to encode the secret, making it much more stable in the turbulent city air.

2. The Method: Discrete "Stepping Stones" vs. a Smooth Ramp

Usually, to send a message, you might imagine sliding a dial smoothly from 0 to 100. This is called "continuous modulation."

• The Issue: Real computers and electronics aren't perfect; they can't make infinite tiny steps. They can only click to specific numbers (like 1, 2, 3, 4).
• The Innovation: Instead of trying to force a smooth dial onto a digital computer, this team embraced the "clicks." They used Discrete Modulation. Think of it like a digital clock that only shows 12, 3, 6, and 9 (a 4-step system, or QPSK). It's easier to build with real hardware and less prone to errors.

3. The Security: The "No-Guessing" Proof

In the past, scientists had to make a big assumption to prove their system was safe: "We assume the spy uses the most common, boring type of attack." This is like saying, "We are safe because we assume the burglar only uses a skeleton key, not a laser cutter."

• The Breakthrough: This paper introduces a new Security Proof that doesn't make that assumption. It's like checking the house for every possible way a burglar could get in, not just the skeleton key.
• Finite Size: Most previous tests only worked if you sent infinite messages. But in real life, you stop after a while. This team proved their system is safe even for a finite (limited) number of messages, which is what you actually need for a real application.

4. The Experiment: The "Honest" Lab Test

They built a mobile lab (a "breadboard" system) with a sender (Alice) and a receiver (Bob).

• The Setup: They didn't use a real windy day yet. Instead, they put a filter in the middle to simulate the loss of signal you'd get in the air.
• The Process:
  1. Alice sends a stream of light pulses with the 4-step "shake" pattern.
  2. Bob catches them and measures the "shake."
  3. The Pipeline: They ran the data through a full software suite (called AIT-QPS). This included:
     • Error Correction: Like two people reading a noisy phone call back and forth to fix typos.
     • Privacy Amplification: Like taking a long, messy list of numbers and compressing it into a short, unbreakable password that Eve can't guess.

5. The Big Discovery: The "Frame Error" Trap

This is the most practical part of their finding.

• The Trap: In the past, researchers thought: "If our error correction is 95% efficient, we just multiply our speed by 0.95." They assumed that if a few messages failed to correct, you just threw them away and kept going.
• The Reality: The team found that if you push the system too hard to get high speed, the error correction starts failing on whole blocks of data (called Frame Errors).
• The Analogy: Imagine you are trying to solve a puzzle. If you rush, you might get 99% of the pieces right, but if you get one whole section wrong, you have to throw away the entire puzzle and start over.
• The Result: They found that trying to be too efficient actually reduced the final secret key size. By slowing down slightly to ensure the error correction didn't fail, they generated a 1.6 Megabyte secret key with a security guarantee so high that the chance of a hack is less than 1 in 10 billion ($10^{-10}$).

Summary

The authors built a "quantum walkie-talkie" designed for the messy, windy air of a city. They proved it's safe without making lazy assumptions about the enemy. Most importantly, they showed that in the real world, slowing down to avoid mistakes is better than rushing and losing your secret key. They successfully generated a large, unbreakable secret key in a lab, proving this technology is ready to move beyond fiber cables and into the open sky.

Technical Summary

Technical Summary: Composable Free-Space Continuous-Variable Quantum Key Distribution Using Discrete Modulation

Problem Statement
Continuous-variable (CV) quantum key distribution (QKD) offers a pathway to quantum-secure communication compatible with existing classical coherent infrastructure. While Gaussian-modulated protocols have been extensively studied and demonstrated in fiber networks, they rely on a theoretical idealization of continuous distributions. Practical implementations are constrained by the finite digital resolution of electronic devices, necessitating discrete-modulated (DM) systems.

A critical gap exists in the experimental realization of DM-CV QKD with rigorous security guarantees. Previous experimental demonstrations of DM-CV QKD have largely been limited to calculating asymptotic key rates, often assuming idealized error correction efficiencies or ignoring the impact of frame errors (FER) on the final secret key. Furthermore, existing security proofs for discrete modulation often rely on Gaussian assumptions or lack composable security in the finite-size regime, which is a prerequisite for cryptographic applications. Additionally, while fiber-based systems are mature, expanding QKD to free-space urban atmospheric channels requires specific adaptations to handle turbulence without the birefringence issues found in other encoding schemes.

Methodology
The authors present a mobile CV QKD system designed for urban atmospheric channels, employing polarization encoding and discrete Quadrature Phase Shift Keying (QPSK) modulation.

• Physical Layer: The system utilizes the non-birefringent nature of the atmosphere by encoding quantum states in the Stokes parameters of orthogonal circular polarization modes (left- and right-handed). A bright local oscillator (LO) and discrete signal states are prepared in the same spatial mode but orthogonal polarizations. This setup is equivalent to homodyne detection with a transmitted LO, where the LO and signal share the same wavefront distortions, allowing for auto-compensation of turbulence during Stokes detection.
• Signal Preparation: The system operates in burst mode, transmitting frames containing reference, guard, signal, and vacuum states. Signal states are encoded using over-sampled first-order Hermite-Gaussian pulses. This temporal mode choice shifts the signal spectrum away from the low-frequency regime, mitigating low-frequency noise and drifts common in atmospheric links.
• Security Proof: The authors implement a novel security proof based on Ref. [30], which allows for the calculation of composable finite-size key rates against independent and identically distributed (i.i.d.) collective attacks without any Gaussian assumptions. The proof involves:
  • Energy Test: To bound the weight of quantum states outside a finite-dimensional cutoff space.
  • Acceptance Test: To verify that observed statistics fall within a pre-defined set of accepted density operators.
  • Optimization: A non-linear semi-definite program (SDP) is solved to minimize the conditional von Neumann entropy $H(X|E)$ over the set of states passing the tests, utilizing a two-step linearization and duality approach.
• Post-Processing Pipeline: A complete QKD pipeline (AIT-QPS) was implemented, including:
  • Quantum Random Number Generation (QRNG): Based on balanced homodyne detection of the vacuum state.
  • Error Correction: Utilizing Low-Density Parity-Check (LDPC) codes in a reverse reconciliation mode.
  • Privacy Amplification: Using Toeplitz matrix hashing.
  • Frame Error Handling: Unlike previous approaches that apply a simple scaling factor $(1-\text{FER})$ to the key rate, this implementation strictly adheres to the security statement. Failed blocks are fully disclosed, increasing the leakage term ($\text{leak}_{EC}$) in the key rate formula, which directly reduces the extractable key length.

Key Contributions

1. First Experimental Demonstration of Composable Finite-Size DM-CV QKD: The paper reports the first laboratory demonstration of a CV QKD system using discrete modulation that calculates composable finite-size key rates against i.i.d. collective attacks without Gaussian assumptions.
2. Polarization Encoding for Free-Space: The system successfully demonstrates polarization encoding in a turbulent, non-birefringent atmospheric channel, offering a viable alternative to amplitude/phase encoding for free-space links.
3. Rigorous Treatment of Frame Errors: The work highlights the critical impact of frame errors on secret key generation. By treating failed error correction blocks as fully disclosed information rather than simply scaling the key rate, the authors demonstrate that operating too close to the error correction threshold (to maximize efficiency) can result in zero or significantly reduced secret keys due to the high cost of disclosing failed blocks.
4. Full Protocol Implementation: The experiment integrates the entire QKD stack, from QRNG and state preparation to error correction and privacy amplification, achieving a total security parameter of $\epsilon < 1 \cdot 10^{-10}$.

Results
In a laboratory demonstration emulating a 3 dB loss channel (representing an atmospheric link), the system achieved the following:

• Key Rates: Composable finite-size key rates of up to $2.26 \cdot 10^{-2}$ bits per symbol were achieved against collective i.i.d. attacks.
• Secret Key Generation: For a block size of $N = 1.2 \cdot 10^9$ states, the system successfully extracted secret keys with a maximum length of 1.6 MB.
• Security Parameters: The total security parameter was bounded at $\epsilon_{imp} < 1 \cdot 10^{-10}$.
• Impact of Frame Errors: The study compared different operating points. While operating near the threshold of the LDPC codes yielded high error correction efficiency ($\beta \approx 89\%$), the resulting Frame Error Rate (FER) of ~16% led to a secret key of only 0.19 MB. By moderating the amplitude to reduce the FER to ~5.9% (while maintaining $\beta \approx 88\%$), the extracted secret key increased to 1.6 MB. This confirms that minimizing FER is crucial for maximizing the final key length when using the rigorous security framework.

Significance
The paper claims that this work bridges the gap between theoretical security proofs for discrete-modulated CV QKD and practical cryptographic applications. By demonstrating a system that generates composable finite-size keys without Gaussian assumptions and by rigorously accounting for frame errors in the security analysis, the authors provide a robust foundation for deploying CV QKD in real-world urban atmospheric networks. The results underscore that for practical applications, optimizing for the highest asymptotic efficiency is insufficient; the system must be tuned to minimize frame errors to ensure the generation of a non-zero, secure secret key. This approach paves the way for expanding QKD networks beyond fiber backbones into free-space environments.