⚛️ quantum physics

No advantage from classical assistance for energy-constrained bosonic dephasing channels

This paper proves that for single-mode bosonic dephasing channels under an average photon-number constraint, interactive classical assistance provides no advantage, as the unassisted and classically assisted quantum and private capacities all coincide with the coherent information optimized over Fock-diagonal input states.

Original authors: Amir Arqand, Graeme Smith

Published 2026-10-12

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Amir Arqand, Graeme Smith

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

In the quiet, controlled world of quantum physics, scientists are constantly trying to figure out how to send information across distances without it getting scrambled by the environment. Imagine trying to whisper a secret across a noisy room; the challenge is to keep the message clear despite the static. In the quantum realm, this static comes from two main sources: the loss of energy, like a signal fading away, and the loss of phase, which is a subtle kind of confusion that happens when the delicate timing of a particle's wave gets jumbled. These two problems affect the very light and microwave signals used to carry quantum data. Researchers have long known that if you are allowed to use a two-way conversation—sending messages back and forth to correct mistakes—you can sometimes overcome these errors much better than if you are forced to send a message in just one direction. This is particularly true for signals that lose energy, where a little bit of back-and-forth talk can turn a broken connection into a working one. However, there is a catch: real-world devices cannot use infinite amounts of energy. They are limited by how much power they can safely handle, which translates to a limit on the average number of light particles, or photons, they can use at any given time.

The question that Amir Arqand and Graeme Smith set out to answer is whether this helpful back-and-forth conversation still provides an advantage when the energy is strictly limited, specifically for the type of noise that scrambles phase rather than losing energy. They focused on a specific kind of communication channel where the signal stays the same strength but gets confused in its timing. In many other scenarios, allowing the sender and receiver to chat and adjust their strategy based on previous results would boost the speed at which they can send secret keys or quantum data. But for this specific type of noisy channel, the researchers proved that the extra conversation does not help at all. They demonstrated that the maximum speed at which you can send information, whether you are allowed to chat or not, is exactly the same, provided you stay within the energy limit.

The team showed that this limit is determined by a single, specific calculation involving the distribution of the light particles used to send the message. They found that no matter how cleverly the parties try to adapt their strategy, using past results to choose future inputs, they cannot beat the rate achieved by a simple, non-adaptive approach. This result is significant because it closes a gap in our understanding of quantum communication. It tells us that for this particular type of noise, the complex machinery of interactive protocols offers no extra power when energy is constrained. The proof relies on comparing the messy, noisy output of the real channel against a perfectly scrambled version of the same signal. By measuring the difference between these two using a mathematical tool called relative entropy, the authors established a strict upper bound on how much information can be preserved. They showed that even if the sender and receiver share complex quantum memories and exchange unlimited classical messages, the rate of information transfer cannot exceed the value calculated for a simple, one-way transmission.

The researchers also confirmed that this limit is not just a theoretical ceiling that cannot be reached; it is a real, achievable speed. They found that by using specific types of light states—those where the number of photons is well-defined but the timing is randomized in a particular way—the parties can hit this maximum rate exactly. This means that for engineers building future quantum networks that deal with phase noise, there is no need to design overly complex systems that rely on constant feedback loops to improve performance. The most efficient strategy is already known, and it does not require the parties to talk to each other during the transmission. The study provides a definitive answer for a wide range of energy limits, from very low power up to any finite amount, ensuring that the rules of quantum communication are clear for this specific, common type of noise.

In the end, the work clarifies the landscape of quantum communication by showing that not all noise is created equal. While some types of errors can be fixed with a little help from a two-way conversation, others are so fundamental that no amount of chatting can improve the situation beyond a certain point. For the specific case of phase noise in light-based systems, the limit is fixed, and the path to the best performance is straightforward. This insight helps strip away unnecessary complexity from future designs, allowing scientists to focus on other challenges where interaction might still make a difference. The findings stand as a rigorous proof, leaving no room for doubt that in this specific scenario, the extra effort of adaptive protocols yields no return on investment.

Technical Summary: No Advantage from Classical Assistance for Energy-Constrained Bosonic Dephasing Channels

Problem Statement
The paper addresses the fundamental question of whether two-way classical communication (LOCC) and adaptive protocols can increase the quantum and private capacities of bosonic dephasing channels when subject to an average photon-number constraint. While previous work established that for unrestricted bosonic dephasing channels, the assisted and unassisted capacities coincide, it remained an open question whether this equality holds under finite energy constraints. Specifically, it was unclear if adaptive protocols, which allow parties to exchange classical messages and adjust inputs based on previous outcomes, could exceed the unassisted rate while maintaining a fixed average input energy $E$ . This is distinct from pure-loss channels, where assistance significantly boosts capacity, and quantum-limited amplifiers, where it does not.

Methodology
The authors employ a framework based on relative entropy comparisons and chain rules to derive upper bounds on the achievable rates of adaptive protocols. The core methodology involves:

Relative Entropy Comparison: The authors compare the output of the bosonic dephasing channel $\mathcal{N}_\mu$ with the output of a completely dephasing channel $\Delta$ . They establish a bound showing that the relative entropy between the channel output and the completely dephased output is bounded by the coherent information $I_c(\rho_A)$ of the input state.
Pinching Identity and Chain Rules: By combining the relative entropy bound with a pinching identity (Lemma 3.2), the authors derive "chain rules" for two key entanglement measures: the Rains relative entropy ( $R$ ) and the relative entropy of entanglement ( $E_R$ ). These chain rules demonstrate that the increase in these measures across a single channel use is bounded by the coherent information of the input marginal, even when the input is correlated with local memories.
Converse Proof for Adaptive Protocols: The authors apply these chain rules iteratively to an $n$ -use adaptive protocol. By leveraging the monotonicity of $R$ and $E_R$ under LOCC operations and the concavity of the coherent information, they show that the total increase in these entanglement measures over $n$ uses is bounded by the sum of the coherent informations of the individual input states.
Energy Constraint Handling: The analysis explicitly accounts for the average energy constraint $E$ imposed on the mean photon number averaged over all channel uses. The proof utilizes bounds on the von Neumann entropy of states with bounded mean photon number (Lemma 2.3) to ensure all terms remain finite.

Key Contributions and Results
The primary result of the paper is the proof that for a single-mode bosonic dephasing channel $\mathcal{N}_\mu$ and any finite energy constraint $E$ , four distinct capacity measures coincide:

The unassisted quantum capacity, $Q(\mathcal{N}_\mu, E)$ .
The unassisted private capacity, $P(\mathcal{N}_\mu, E)$ .
The two-way classically assisted quantum capacity, $Q_{\leftrightarrow}(\mathcal{N}_\mu, E)$ .
The two-way public-classically assisted private capacity, $P_{\leftrightarrow}(\mathcal{N}_\mu, E)$ .

All four capacities are given by the same single-letter formula:
$Q_{\leftrightarrow} = P_{\leftrightarrow} = Q = P = \sup_{\rho(k)} I_c(\rho_A)$
where the supremum is taken over all Fock-diagonal input states $\rho_A = \sum_k \rho(k)|k\rangle\langle k|$ satisfying the average photon-number constraint $\sum_k k\rho(k) \leq E$ . The coherent information is defined as $I_c(\rho_A) = H(\rho_A) - H(\sum_k \rho(k)|e_k\rangle\langle e_k|_F)$ , where $|e_k\rangle_F$ are the environment states associated with the Fock inputs.

The paper further establishes that the derived upper bounds on the Rains relative entropy and the relative entropy of entanglement are tight. Specifically, these bounds are attained by the canonical purifications of finite-energy Fock-diagonal states (Corollary 3.1). The proof holds for protocols with countably supported photon-number distributions and does not require finite-dimensional restrictions on local memories or private-state shields.

Significance
The paper concludes that for bosonic dephasing channels, adaptive classical assistance provides no advantage over unassisted protocols when the communication is constrained by a finite average energy. This result extends the known equality of assisted and unassisted capacities from the energy-unrestricted regime to the physically relevant energy-constrained regime. The work provides a complete single-letter characterization of these capacities, resolving the question of whether adaptive strategies can overcome the limitations imposed by the energy constraint in dephasing environments. The authors note that while the characterization is complete in terms of optimization over photon-number distributions, identifying the specific optimal distributions for given phase-noise models remains an open direction. Additionally, the paper suggests that the developed methods could be applied to study bosonic loss-dephasing channels, which combine photon loss and dephasing.

Technical Summary: No Advantage from Classical Assistance for Energy-Constrained Bosonic Dephasing Channels

More like this