On capacity of optical communications over a lossy bosonic channel with a receiver employing the most general coherent electro-optic feedback control.

Hye Won Chung, Saikat Guha, Lizhong Zheng

TL;DR

This paper investigates the capacity of optical communication channels using a specific type of receiver that employs coherent electro-optic feedback. The authors generalize a known optimal receiver for distinguishing two coherent states to multiple states and analyze its performance. Surprisingly, they find that this sophisticated receiver, even with feedback, doesn't offer an asymptotic advantage in communication rate over a simple direct-detection receiver when using long codewords.

What Problem Does This Paper Address?

Optical communication systems often rely on transmitting information encoded in coherent states of light. A key challenge is reliably distinguishing these states at the receiver, especially in the presence of noise and loss. While direct detection (simply counting photons) is a standard approach, more advanced receivers can use coherent feedback to process the incoming signal before detection. The paper builds upon Dolinar's work, which showed that a specific feedback receiver can achieve the theoretical minimum error probability for distinguishing two coherent states. The authors aim to generalize this to discriminating multiple coherent states and to understand the ultimate communication capacity achievable by such receivers, particularly in the low-photon regime where quantum effects are significant. They want to know if complex feedback mechanisms can truly boost the rate beyond what's possible with simpler methods and good coding.

Methods

The core of the paper involves generalizing Dolinar's optimal receiver for binary hypothesis testing to the M-ary case. This is done by re-interpreting Dolinar's receiver as one that recursively optimizes information efficiency at each time step using Bayesian updates. For M-ary discrimination, they explore using Rényi entropies as a metric to guide the control signal optimization, aiming to minimize conditional entropy. They then analyze the channel capacity achievable by this generalized receiver, focusing on the low-photon number regime. A key technique is deriving theoretical bounds on the achievable information rate and comparing them to the Holevo limit (the theoretical maximum capacity) and the capacity of a direct-detection receiver. They employ mathematical induction and Fano's inequality to establish these bounds.

Datasets

Not discussed in the paper. The paper focuses on theoretical analysis and performance bounds rather than empirical evaluation on specific datasets.

Key Insights and Ideas

  • Generalizing Dolinar's Receiver: The paper provides a new perspective on Dolinar's receiver by framing it as an incremental optimization of information efficiency, which naturally leads to a generalization for M-ary hypothesis testing.
  • Rényi Entropy for Optimization: The use of Rényi entropies as a metric for optimizing the feedback control signal is a novel approach for M-ary discrimination, allowing for adaptive strategies based on the remaining time for decision-making.
  • The "No-Go" Theorem for Capacity: The most striking finding is that, in the limit of infinitely long codewords, the asymptotic communication rate achievable by the generalized coherent-processing receiver is no better than that of a simple direct-detection receiver. This implies that for maximizing rate, the benefits of coherent feedback are superseded by the power of coding.
  • Implications for Quantum vs. Classical Processing: The results suggest that to surpass the capacity of direct detection and approach the Holevo limit, truly quantum processing (e.g., using entangled states or squeezed light) within the receiver might be necessary, rather than just classical coherent feedback.
  • Practical Implications for Low-Photon Regimes: While the asymptotic rate is the same, the paper highlights that for photon-starved scenarios, the Holevo capacity receiver could offer significantly higher data rates than direct detection, even if the generalized coherent receiver doesn't bridge this gap.

Key Results

  • The paper derives a generalized receiver structure for M-ary coherent-state discrimination based on recursive mutual information maximization.
  • Theorem 4: The achievable photon information efficiency for a coherent-processing receiver (as described) is bounded above by $E \log(1/\mathcal{E}) - E \log \log(1/\mathcal{E}) + O(\mathcal{E})$, where $\mathcal{E}$ is the mean photon number. This is the same asymptotic scaling as that of a direct-detection receiver (Lemma 3).
  • This implies that for long codewords, the coherent-processing receiver offers no asymptotic advantage in communication rate over direct detection.
  • The Holevo capacity, achievable with ideal joint quantum detection, is shown to have a better second-order term, indicating a significant advantage in photon-starved regimes.

Limitations and Future Work

  • Conjecture on General Receivers: The authors conjecture that their "no-go" result extends to even more general passive linear-optical pre-processing circuits before the feedback receiver, but this remains unproven.
  • Focus on Coherent States: The analysis is primarily for coherent states. Extending the results to other quantum states (e.g., squeezed states) would require different theoretical tools.
  • Optimal Time-Varying Rényi Parameter: The paper suggests that an optimal strategy might involve a time-varying parameter $\alpha$ in the Rényi entropy, which is an interesting avenue for future work.
  • Non-Classical Control Signals: Investigating receivers that use non-classical states (like squeezed states) for local control signals is proposed as a path to potentially achieve higher capacities.
  • Experimental Verification: While the paper presents theoretical results, experimental validation of these findings would be a crucial next step.

Summary updated

Abstract

We study the problem of designing optical receivers to discriminate between multiple coherent states using coherent processing receivers---i.e., one that uses arbitrary coherent feedback control and quantum-noise-limited direct detection---which was shown by Dolinar to achieve the minimum error probability in discriminating any two coherent states. We first derive and re-interpret Dolinar's binary-hypothesis minimum-probability-of-error receiver as the one that optimizes the information efficiency at each time instant, based on recursive Bayesian updates within the receiver. Using this viewpoint, we propose a natural generalization of Dolinar's receiver design to discriminate $M$ coherent states each of which could now be a codeword, i.e., a sequence of $N$ coherent states each drawn from a modulation alphabet. We analyze the channel capacity of the pure-loss optical channel with a general coherent-processing receiver in the low-photon number regime and compare it with the capacity achievable with direct detection and the Holevo limit (achieving the latter would require a quantum joint-detection receiver). We show compelling evidence that despite the optimal performance of Dolinar's receiver for the binary coherent-state hypothesis test (either in error probability or mutual information), the asymptotic communication rate achievable by such a coherent-processing receiver is only as good as direct detection. This suggests that in the infinitely-long codeword limit, all potential benefits of coherent processing at the receiver can be obtained by designing a good code and direct detection, with no feedback within the receiver.
BibTeX
@inproceedings{chung2017capacity,
  author = {Hye Won Chung and Saikat Guha and Lizhong Zheng},
  title = {{On capacity of optical communications over a lossy bosonic channel with a receiver employing the most general coherent electro-optic feedback control.}},
  booktitle = {Physical},
  year = {2017},
  doi = {10.1103/physreva.96.012320},
}
Figure