← Latest papers
🔢 mathematics

Entanglement cost and classical capacity of multimode Gaussian states and channels

This paper establishes that for finite-energy bipartite Gaussian states, the entanglement of formation, Gaussian entanglement of formation, and entanglement cost coincide and are strongly superadditive, while also determining the exact classical capacity and entanglement cost of multimode Gaussian channels under mean energy constraints through Gaussian covariance optimizations.

Original author: Yuan Gao, Gerardo Adesso

Published 2026-11-18
📖 6 min read🧠 Deep dive

Original author: Yuan Gao, Gerardo Adesso

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 realm of quantum physics, there is a special class of states known as Gaussian states. These are not exotic, fragile creations that exist only in the most extreme laboratory conditions; rather, they are the workhorses of continuous-variable quantum information, the kind of systems where information is carried by the smooth, wave-like properties of light and other bosonic systems. Think of them as the standard, reliable tools for encoding data in the vibrations of light beams. Because these states are mathematically well-behaved and can be described completely by their average values and how much they fluctuate, they form the backbone of modern quantum communication networks. However, a fundamental question has lingered for years: when we need to create these states or send information through them, is the most efficient way always to use these simple, Gaussian tools, or do we need to resort to more complex, irregular quantum states to save resources?

A team of researchers has now answered this question with a definitive proof. They demonstrated that for any system involving a finite number of light modes, the most efficient way to prepare a shared quantum state is always to use a Gaussian method. Specifically, they showed that the amount of entanglement required to create a Gaussian state is exactly the same whether you try to build it using only simple Gaussian building blocks or if you allow yourself to use any possible quantum trick, no matter how complex. This means that for these systems, the simple approach is not just a good approximation; it is the absolute best possible strategy. Furthermore, they calculated exactly how much entanglement is needed to simulate any type of noisy quantum channel and determined the maximum rate at which classical information can be sent through such channels when energy is limited. Their findings reveal that the optimal way to communicate is to use pure Gaussian signals that are slightly shifted in a random, Gaussian pattern, and that this strategy works perfectly even when the channels are complex and the noise is correlated across different parts of the system.

The researchers focused on a concept called entanglement cost, which is essentially a measure of how much "quantum fuel" is required to create a specific shared state between two parties. In the quantum world, creating a shared state often requires consuming entanglement, a resource where two particles remain linked regardless of distance. For years, scientists debated whether using complex, non-Gaussian states could reduce this cost. The new work proves that for any finite-energy Gaussian state, the cost is identical whether you restrict yourself to Gaussian preparations or allow any quantum method. This result holds true for systems with any number of modes, from a single pair of light beams to complex networks involving many. The team also showed that this cost is "strongly superadditive," meaning that if you have a large system made of several smaller groups, the total entanglement required to create the whole system is at least the sum of the costs for the individual groups. This property ensures that you cannot save resources by trying to prepare many copies of a state all at once; the cost per copy remains constant.

Beyond just creating states, the paper tackles the problem of simulating quantum channels, which are the pathways that carry information. The researchers calculated the exact entanglement cost required to simulate any multimode Gaussian channel, including those with correlated noise and unequal losses. They found that this cost is determined by how much the channel distorts the quantum state, and it can be calculated by looking at the state produced when the channel acts on a specific type of entangled probe. If the channel adds no noise to certain parts of the signal, the cost becomes infinite, meaning it is impossible to simulate perfectly with finite resources. However, if the channel adds enough noise, a finite amount of entanglement is sufficient. The team provided a clear formula for this cost, showing that it depends only on the mathematical description of the channel's noise and attenuation.

The study also addressed the classical capacity of these channels, which is the maximum speed at which classical information can be transmitted. By connecting the problem of sending information to the problem of creating entanglement, the researchers proved that the optimal strategy for communication involves sending pure Gaussian signals that are modulated by random shifts. They showed that this strategy achieves the highest possible data rate under a given energy constraint. Interestingly, they found that for channels with unequal losses and correlated noise, using an entangled Gaussian signal across the input ports can provide a strictly higher data rate than using independent signals for each port. This advantage comes from the ability to exploit the correlations within the channel itself, a feat that is impossible if one treats the channels as separate, independent lines.

To make these theoretical results practical, the authors developed a numerical toolkit that allows researchers to calculate these entanglement costs and capacities for any specific setup. This software can handle complex, real-world scenarios where symmetries are broken, providing upper and lower bounds that are often very close to the exact value. The team also demonstrated that their findings apply to non-Gaussian states as well. They showed that if you take an optimal Gaussian preparation and replace the random shifts with a different, non-Gaussian distribution that has the same average properties, the entanglement cost remains exactly the same. This means that even for systems that are not perfectly Gaussian, the simple Gaussian formulas provide a strict lower bound on the resources needed, allowing scientists to estimate the difficulty of preparing or communicating with these states without needing to know every detail of the system.

The implications of this work extend to the design of future quantum networks. By proving that Gaussian strategies are optimal for a wide range of tasks, the researchers have removed the need to search for more complex, potentially more expensive solutions. Their results provide a solid foundation for designing communication protocols that make the most efficient use of quantum resources. The ability to calculate exact costs and capacities for correlated, multimode channels means that engineers can now design systems that are tailored to the specific noise and loss characteristics of their hardware, rather than relying on approximations. As experimental techniques improve and larger, more complex optical systems are built, these formulas will serve as the blueprint for maximizing the performance of quantum communication, ensuring that the transmission of information is as efficient as the laws of physics allow.

Drowning in papers in your field?

Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.

Try Digest →