Pith. sign in

REVIEW 3 major objections 6 minor 29 references

Data Transmission over a Bosonic Arbitrarily Varying Quantum Channel

T0 review · 3 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper establishes a min-max capacity formula for classical data transmission over a lossy bosonic channel with a semi-classical jammer, and derives a closed-form expression conditional on a conjectured entropy power inequality.

desk verdict Solid technical work on bosonic AVC capacity, but Theorem 4's converse is a one-line placeholder, leaving the central claim unproven until the missing minimax step is provided. read the letter →

arxiv 2507.18259 v1 pith:G3743QPL submitted 2025-07-24 quant-ph

Janis Nötzel , Florian Seitz This is my paper
classification quant-ph MSC 81P4594A40
keywords arbitrarilyvaryingchannelbosonicbeam-splitterquantumjammingcommonrandomnessHolevoinformationentropypowerinequalitycapacityformula
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper asks how much classical data can be sent reliably over a lossy bosonic channel when an adversary injects semi-classical noise into the same transmission line, a model for free-space optical links and non-amplified fiber under jamming. It proves that, with common randomness shared by sender and receiver, the capacity is $\sup_{\mu \in \mathcal{D}_E} \inf_{\sigma \in \mathcal{S}_P} \chi(\mu; \mathcal{N}_\sigma)$: the sender picks an input ensemble to maximize the worst-case Holevo information against the allowed jammer states. If a recently conjectured quantum entropy power inequality is true, this min-max expression collapses to $g(\tau E + \tau' P) - g(\tau' P)$, depending only on the beam-splitter transmissivity and the two energy budgets. The result matters because it turns an adversarial infinite-dimensional problem into a finite optimization with a simple physical reading: the jammer's best strategy is a thermal state.

What carries the argument

The machine is the beam-splitter channel $\boxplus_\tau$, where the sender's coherent state and the jammer's state are mixed and one output port is observed. The proof's engine is symmetrization by common randomness: random phase rotations and permutations convert an arbitrary allowed jamming strategy into a permutation-invariant mixture of i.i.d. attacks, so standard compound-channel coding applies after a de-Finetti-type estimate; the effective local dimension grows only as $\log k$, which keeps the truncation error small. For the closed form, the load-bearing identity is the conjectured entropy power inequality (C), $S(X \boxplus_\lambda Y) \ge g(L_\lambda(X)+R_\lambda(Y))$, where $g(x)=(x+1)\log(x+1)-x\log x$ and $L_\lambda,R_\lambda$ are defined through $g^{-1}\circ S$ of the beam-split output against vacuum. This inequality reduces the min-max to a one-line optimization whose optimum is the thermal state of mean photon number $P$.

What would settle it

Find two one-mode bosonic states $X,Y$ and a splitting ratio $\lambda$ with $S(X \boxplus_\lambda Y) < g(L_\lambda(X)+R_\lambda(Y))$; a single such counterexample would disprove Conjecture (C) and invalidate Theorem 5 while leaving Theorem 4 intact. Separately, test whether a heavy-tailed, non-sub-Gaussian jammer distribution changes $\inf_{\sigma\in\mathcal{S}_P}\chi(\mu;\mathcal{N}_\sigma)$; if it does, the sub-Gaussian restriction is not merely technical.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central claim is a coding theorem for the beam-splitter channel $\boxplus_\tau$: with coherent-state inputs, jammer states with positive $P$-representation (mixtures of coherent states), and average energy constraints $E$ (sender) and $P$ (jammer), the common-randomness-assisted classical capacity is $C(\mathcal{N}) = \sup_{\mu\in\mathcal{D}_E}\inf_{\sigma\in\mathcal{S}_P}\chi(\mu;\mathcal{N}_\sigma)$, and rates up to this value are achievable with random code ensembles of support size at most $k^2$. Granting Conjecture (C), $S(X \boxplus_\lambda Y) \ge g(L_\lambda(X)+R_\lambda(Y))$, the same capacity equals $g(\tau E + \tau' P) - g(\tau' P)$. The proof symmetrizes any allowed attack by random phase rotations and permutations, turning it into an i.i.d. compound attack, then truncates each mode to dimension $O(\log k)$ and applies a de-Finetti-type estimate together with compound-channel coding lemmas.

Load-bearing premise

The load-bearing premise is the conjectured entropy power inequality $S(X \boxplus_\lambda Y) \ge g(L_\lambda(X)+R_\lambda(Y))$; if it fails, the explicit closed form is unsupported and only the min-max formula of Theorem 4 is proven, and the proof also assumes the jammer's distribution is sub-Gaussian rather than allowing arbitrary heavy tails.

Editorial extensions

If this is right

  • Rates arbitrarily close to the min-max capacity are achievable with common randomness using random code ensembles whose support grows only as $k^2$.
  • If Conjecture (C) holds, the capacity of the lossy bosonic channel under semi-classical jamming is exactly $g(\tau E+\tau' P)-g(\tau' P)$ and does not depend on the fine structure of the signal ensemble.
  • The jammer's optimal strategy is a thermal state with mean photon number $P$, so only the jammer's energy budget enters the capacity.
  • The classical Gaussian arbitrarily varying channel capacity picture transfers to bosonic quantum channels, with the entropy function $g$ replacing the logarithm.
  • The $O(\log k)$ effective dimension means the coding theorem works through a finite-dimensional truncation whose cost grows only logarithmically with block length.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper, a proof of Conjecture (C) would immediately turn the capacity result into a finite-blocklength program, because the truncation dimension grows only logarithmically.
  • Beyond the paper, the sub-Gaussian restriction on the jammer is likely removable or relaxable; testing whether heavy-tailed jammers change the min-max value is a direct next step.
  • Beyond the paper, a multimode version of the capacity formula, with total energies $E$ and $P$ distributed across modes, is a natural extension that the single-mode proof structure seems to support.
  • Beyond the paper, the formula gives a concrete experimental benchmark: on a beam-splitter link with known loss, the predicted rate under worst-case semi-classical jamming can be compared with the expression in a laboratory.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper considers a bosonic beam-splitter channel ⊞_τ in which a legitimate sender controls one input port and an adversary (jammer) controls the other, with the receiver observing one output port. The jammer is restricted to semi-classical (positive P-representation) states, and both sender and jammer are subject to energy constraints E and P, respectively, with an additional sub-Gaussianity constraint on the jammer's P-function. The main result, Theorem 4, states that the common-randomness-assisted capacity equals C(N) = sup_{μ∈D_E} inf_{σ∈S_P} χ(μ; N_σ), where S_P is the set of admissible single-use jamming states and D_E is the sender's input distribution set (though D_E is not explicitly defined). The proof of achievability uses phase randomization, a de Finetti-type reduction adapted to energy constraints, and compound-channel coding results. The paper also gives Theorem 5, a closed-form expression C(N) = g(τE+τ'P)−g(τ'P), conditional on a conjectured entropy power inequality (Conjecture C) from the authors' earlier work [16]. The conclusion acknowledges that the sub-Gaussianity condition is a technical restriction.

Significance. If fully established, Theorem 4 would be the first coding theorem for a bosonic arbitrarily varying channel with adversarial semi-classical jamming, extending classical Gaussian AVC results and potentially informing free-space optical communication security. The achievability proof is a substantial technical contribution: it combines phase randomization with a dimension-truncation argument (N ∼ log k), a constrained de Finetti bound, and a reduction to compound-channel coding, and it correctly identifies the conjectured entropy power inequality as the missing tool for an explicit formula. The paper is transparent about the conditional nature of Theorem 5 and about the technical sub-Gaussian restriction, which is commendable. However, the current manuscript lacks a rigorous converse for Theorem 4, and there is a definitional mismatch between the jammer set used in the code definition and the set used in the capacity formula; both issues are load-bearing for the paper's central claim.

major comments (3)
  1. [Section V, last paragraph] The converse of Theorem 4 is not proven. The sentence 'The converse follows by assuming a jamming strategy tailored to (IV.1), where the jammer sends the jamming signal S_P^{\otimes k}' is insufficient for two reasons. First, S_P^{\otimes k}, the tensor power of the thermal state, is not an element of S_P^k as defined in Section II, because S_P^k consists of product coherent states ⊗_{i=1}^k |α_i⟩⟨α_i| with sub-Gaussian empirical distribution. Second, even if S_P^{\otimes k} were allowed, it would only give the upper bound C(N) ≤ sup_{μ∈D_E} χ(μ; N_{S_P}) = g(τE+τ'P)−g(τ'P), which is not the claimed min-max expression sup_{μ∈D_E} inf_{σ∈S_P} χ(μ; N_σ). To prove (IV.1) one must show that for every rate above the min-max value there exists an allowed jammer sequence σ^k ∈ S_P^k such that the average success probability is bounded away from one; this minimax argument over the noncompact sets D_E and S_P is not supplied.
  2. [Section II / Definition 2] There is a definitional mismatch between the jammer set used in the code definition and the set used in the capacity formula. Definition 2 defines the AVC via the infimum over σ ∈ S_P^k, where S_P^k is defined in Section II as the set of product coherent states with sub-Gaussian empirical distribution, whereas the optimization in (IV.1) is over the single-use set S_P of all PHAV states with sub-Gaussian P-function. Since the coherent states form a strict subset of S_P, the infimum in the formula can be strictly smaller than the worst case over the actually allowed product-coherent sequences, so (IV.1) may be a lower bound rather than the exact capacity. The achievability proof can be read as enlarging the jammer set via CR and permutation averaging, but the converse must defeat codes using sequences from the original product-coherent set; the paper does not address this.
  3. [Section VI] The proof of Theorem 5 relies on the assertion 'SP maximizes S(σ) on SP' without checking that the thermal state S_P belongs to the admissible set S_P. Under the definition in Section II, a state is in S_P only if its P-function is sub-Gaussian with constant K1 ≤ P. For a thermal state with mean photon number P the P-function is Gaussian with variance P, whose sub-Gaussian constant is of order √P; hence the condition K1 ≤ P holds only when P is larger than a constant of order one. For small P the thermal state is not admissible, the maximum of R_τ(σ) over S_P is not attained at S_P, and the closed-form expression g(τE+τ'P)−g(τ'P) does not follow from the argument given. This gap is distinct from the unproven Conjecture (C) and requires either a relaxation of the sub-Gaussian condition or a separate treatment of the P < 1 regime.
minor comments (6)
  1. [Section IV] The set D_E used in Theorem 4 is never defined; presumably it is the set of probability distributions on C with mean photon number ≤ E, but this should be stated explicitly.
  2. [Section II] The notation S_F for the set of PHAV states and S_N for the thermal state is confusing, and in Section VI 'SP' is used both for the thermal state and for the set S_P; the paper should use distinct symbols for the set and the thermal state.
  3. [Section V] Equation (V.10) appears to be missing a factor 1/M in the first term after the infimum, and 'MX' should be 'M'.
  4. [Abstract] The abstract's phrase 'we give an explicit capacity formula' should mention that the closed form is conditional on the unproven Conjecture (C), as is done in the introduction.
  5. [Lemma 8] In Lemma 8, the variable α in the condition N ≥ 22·max{2,|α|²} is undefined; the statement should specify that this is the coherent amplitude of the displaced phase-randomized state.
  6. [Conclusion] The conclusion says 'We have derived a capacity formula' but only the min-max expression is unconditional; the analytical formula is conditional on Conjecture (C) and on the sub-Gaussian parameter regime. Please make this distinction explicit.

Circularity Check

2 steps flagged · score 6.0 of 10

Explicit capacity formula is conditional on the authors' own unproved EPI conjecture; the converse of Theorem 4 is asserted by assuming the theorem's min-max jamming strategy.

  1. self citation load bearing [Section IV, Theorem 5; Section VI, Eq. (C) and surrounding derivation]
    "As it turns out, this question can be answered in the affirmative if a recently conjectured [16], new entropy power inequality turns out to be true. ... If the conjectured inequality (C) holds, then the capacity formula in Theorem 4 simplifies to C(N) = g(τ E+ τ ′P ) − g(τ ′P )."

    The paper's advertised closed-form capacity is derived by invoking Conjecture (C), which is unproved and is attributed only to the authors' own prior ISIT paper [16]; Section VI states that the proof will 'closely follow the strategy that was used in [16].' The explicit formula therefore has no independent derivation: it is a consequence of a self-cited conjecture that the present paper does not establish. If (C) fails, Theorem 5 is unsupported. This is a load-bearing self-citation rather than a formal identity, but it is the sole route from Theorem 4 to the advertised formula.

  2. other [Section V-A, final sentence of the proof of Theorem 4]
    "The converse follows by assuming a jamming strategy tailored to (IV.1), where the jammer sends the jamming signal S⊗k P."

    The converse of Theorem 4 must prove the upper bound C(N) ≤ sup_μ inf_σ χ(μ; N_σ). The sentence assumes exactly the object to be established: a 'jamming strategy tailored to (IV.1)' is an adversary choice realizing the infimum in the claimed capacity formula. No argument shows that such a strategy exists for arbitrary codes or that it enforces the bound; naming S_P^{⊗k} gives only the weaker bound sup_μ χ(μ; N_{S_P}). The upper bound is thus imported as an assumption of the theorem's conclusion (or deferred to an unstated minimax theorem), so the converse half of Theorem 4 is circular as written.

full rationale

The achievability half of Theorem 4 is a substantial, independent derivation using phase randomization, de Finetti estimates, and finite-dimensional truncation, and it does not depend on the conjecture. The energy and sub-Gaussian constraints are stated assumptions, and the authors explicitly acknowledge the latter as technical, which is not circular. However, the paper's headline explicit formula is not a first-principles result: Theorem 5 is explicitly conditional on Conjecture (C), an unproved entropy power inequality taken from the authors' own prior work, and the proof of Theorem 5 follows that prior strategy. That is a load-bearing self-citation: the advertised formula is not derived from an external benchmark or an independent theorem. In addition, the one-sentence converse of Theorem 4 is circular as worded, since it assumes a jamming strategy 'tailored to (IV.1)'—the min-max expression whose capacity equality is the theorem being proved. Taking these together, the central derivation chain is partially self-referential: the min-max theorem's converse is assumed rather than shown, and the simplified capacity formula reduces to a conjecture supplied by the same authorship. The score of 6 reflects partial circularity in the central claims, while acknowledging that the achievability argument and the conditional logical structure are honestly stated and carry independent content.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The central claim rests on standard quantum information tools and two assumptions specific to this paper: the unproven entropy power inequality Conjecture (C) and the sub-Gaussianity constraint on jamming. No free parameters are fitted to data; the physical parameters τ, E, P are channel inputs. No new entities are postulated.

assumptions (6)
  • ad hoc to paper Conjectured entropy power inequality: S(X ⊞_λ Y) ≥ g(L_λ(X) + R_λ(Y)) for one-mode bosonic systems (Conjecture (C), Section IV)
    Unproven conjecture from the authors' own prior work [16]; Theorem 5's capacity formula is derived assuming its validity. If false, the closed-form capacity expression does not follow.
  • ad hoc to paper Sub-Gaussianity of the jammer's empirical distribution with constant K1 ≤ F (Section II, definition of S_F)
    Technical restriction imposed to enable de Finetti and concentration arguments. The authors acknowledge in Section VII that this is a restriction to be removed in future work.
  • domain assumption Positive Glauber-Sudarshan P representation for jammer states (Definition 1)
    Defines the semi-classical attack model. Physically motivated but restricts the adversary compared to arbitrary quantum states.
  • domain assumption Energy constraints on sender (E) and jammer (P) via trace Hamiltonian bounds
    Physical power constraints; standard in bosonic channel models.
  • domain assumption Explicit allowance of arbitrary amounts of common randomness (CR) (Definition 2)
    The capacity formula is for CR-assisted codes; the model depends on this resource being available for free.
  • standard math Entropy continuity under energy constraints (Lemma 13 from [22])
    Used to pass from finite-dimensional approximations to infinite-dimensional limits in (V.18)-(V.24).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Data Transmission over a Bosonic Arbitrarily Varying Quantum Channel." pith.science (2026). https://pith.science/paper/G3743QPL

@misc{pith2026250718259,
  author       = {Pith},
  title        = {Pith review of: Data Transmission over a Bosonic Arbitrarily Varying Quantum Channel},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/G3743QPL}},
  note         = {Machine review of arXiv:2507.18259}
}
read the original abstract

Arbitrarily varying channels offer a powerful framework for analyzing the robustness of quantum communication systems, especially for classical-quantum models, where the analysis displays strengths or weaknesses of specific signal constellations under generic attacks. In this work, we provide a coding theorem for a large class of practically relevant arbitrarily varying channel models. Namely, we give an explicit capacity formula for the lossy bosonic channel subject to semi-classical attacks, where an adversary injects semi-classical states into the transmission line. Mathematically, this is modeled via a beam-splitter setup, with transmitter and jammer controlling different input ports and the receiver observing one output port. We show how a recently conjectured new quantum entropy power inequality relates to our capacity formula.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

29 extracted references · 24 canonical work pages

  1. [16]

    Data transmission over a bosonic channel under classical noise,

    J. N ¨otzel, “Data transmission over a bosonic channel under classical noise,” in 2024 IEEE International Symposium on Information Theory (ISIT), 2024, pp. 1227–1232

    work page 2024

  2. [1]

    The capacities of certain channel classes under random coding,

    D. Blackwell, L. Breiman, and A. J. Thomasian, “The capacities of certain channel classes under random coding,” The Annals of Mathematical Statistics, vol. 31, no. 3, pp. 558–567, 1960

    work page 1960

  3. [2]

    Exponential error bounds for random codes in the arbi- trarily varying channel,

    T. Ericson, “Exponential error bounds for random codes in the arbi- trarily varying channel,” IEEE Transactions on Information Theory , vol. 31, no. 1, pp. 42–48, 1985

    work page 1985

  4. [3]

    Elimination of correlation in random codes for arbi- trarily varying channels,

    R. Ahlswede, “Elimination of correlation in random codes for arbi- trarily varying channels,” Z. Wahrscheinlichkeitstheorie verw Gebiete, vol. 44, pp. 159 – 175, 1978

    work page 1978

  5. [4]

    The capacity of the arbitrarily varying channel revisited: positivity, constraints,

    I. Csiszar and P. Narayan, “The capacity of the arbitrarily varying channel revisited: positivity, constraints,” IEEE Transactions on In- formation Theory, vol. 34, no. 2, pp. 181–193, 1988

    work page 1988

  6. [5]

    Classical capacity of classical- quantum arbitrarily varying channels,

    R. Ahlswede and V . Blinovsky, “Classical capacity of classical- quantum arbitrarily varying channels,” IEEE Transactions on Infor- mation Theory, vol. 53, no. 2, pp. 526–533, 2007

    work page 2007

  7. [6]

    Quantum capacity under adversarial quantum noise: Arbitrarily varying quantum channels,

    R. Ahlswede, I. Bjelakovic, H. Boche, and J. N ¨otzel, “Quantum capacity under adversarial quantum noise: Arbitrarily varying quantum channels,” Communications in Mathematical Physics, vol. 317, pp. 103 – 156, 2013. [Online]. Available: https://doi.org/10.1007/s00220-012-1613-x

    doi:10.1007/s00220-012-1613-x 2013

  8. [7]

    Elimination of correlation in random codes for arbi- trarily varying channels,

    R. Ahlswede, “Elimination of correlation in random codes for arbi- trarily varying channels,” Z. Wahrscheinlichkeitstheorie verw. Gebiete, pp. 159–175, 1978

    work page 1978

Show all 29 references
  1. [8]

    Quantum communication with zero-capacity channels,

    G. Smith and J. Yard, “Quantum communication with zero-capacity channels,” Science, vol. 321, no. 5897, pp. 1812–1815, Sep. 2008. [Online]. Available: http://dx.doi.org/10.1126/science.1162242

    2008 doi

  2. [9]

    Super- activation as a unique feature of arbitrarily varying wiretap channels,

    R. F. Schaefer, H. Boche, and H. V . Poor, “Super- activation as a unique feature of arbitrarily varying wiretap channels,” 2016 IEEE International Symposium on Information Theory (ISIT) , pp. 3077–3081, 2016. [Online]. Available: https://api.semanticscholar.org/CorpusID:16496771

    2016

  3. [10]

    N ¨otzel, M

    J. N ¨otzel, M. Wiese, and H. Boche, The Discrete Memoryless Arbi- trarily Varying Wiretap Channel. Cambridge University Press, 2017, pp. 258–312

    2017

  4. [11]

    Capacity of the gaussian arbitrarily varying channel,

    I. Csiszar and P. Narayan, “Capacity of the gaussian arbitrarily varying channel,” IEEE Transactions on Information Theory , vol. 37, no. 1, pp. 18–26, 1991

    1991

  5. [12]

    A generalization of the entropy power inequality to bosonic quantum systems,

    G. De Palma, A. Mari, and V . Giovannetti, “A generalization of the entropy power inequality to bosonic quantum systems,” Nature Photonics, vol. 8, no. 12, pp. 958–964, Nov. 2014. [Online]. Available: http://dx.doi.org/10.1038/nphoton.2014.252

    2014 doi

  6. [13]

    Limits on classical communication from quantum entropy power inequalities,

    R. K ¨onig and G. Smith, “Limits on classical communication from quantum entropy power inequalities,” Nature Photonics , vol. 7, no. 2, pp. 142–146, Jan. 2013. [Online]. Available: http://dx.doi.org/10.1038/nphoton.2012.342

    2013 doi

  7. [14]

    A. S. Holevo, Quantum Systems, Channels, Information . Berlin, Boston: DE GRUYTER, jan 2012

    2012

  8. [15]

    The entropy photon- number inequality and its consequences,

    S. Guha, B. I. Erkmen, and J. H. Shapiro, “The entropy photon- number inequality and its consequences,” 2007

    2007

  9. [17]

    Distinguishing coherent states from phase-mixed coherent states with only a variable beam splitter and single-photon detector, in principle,

    S. R. Hedemann, “Distinguishing coherent states from phase-mixed coherent states with only a variable beam splitter and single-photon detector, in principle,” 2016

    2016

  10. [18]

    Vershynin, High-Dimensional Probability: An Introduction with Applications in Data Science , ser

    R. Vershynin, High-Dimensional Probability: An Introduction with Applications in Data Science , ser. Cambridge Series in Statistical and Probabilistic Mathematics. Cambridge University Press, 2018

    2018

  11. [19]

    A. S. Holevo, Quantum Systems, Channels, Information: A Mathe- matical Introduction . Berlin, Boston: De Gruyter, 06 Dec. 2012. [Online]. Available: https://www.degruyter.com/view/title/122902

    2012

  12. [20]

    Fully quantum arbitrarily varying channels: Random coding capacity and capacity dichotomy,

    H. Boche, C. Deppe, J. N ¨otzel, and A. Winter, “Fully quantum arbitrarily varying channels: Random coding capacity and capacity dichotomy,” in 2018 IEEE International Symposium on Information Theory (ISIT) . IEEE, jun 2018. [Online]. Available: https://doi.org/10.1109

    2018

  13. [21]

    Flexible constrained de finetti reductions and applications,

    C. Lancien and A. Winter, “Flexible constrained de finetti reductions and applications,” Journal of Mathematical Physics , vol. 58, no. 9, Sep. 2017. [Online]. Available: http://dx.doi.org/10.1063/1.5003633

    2017 doi

  14. [22]

    Tight uniform continuity bounds for quantum entropies: conditional entropy, relative entropy distance and energy constraints,

    A. Winter, “Tight uniform continuity bounds for quantum entropies: conditional entropy, relative entropy distance and energy constraints,” Communications in Mathematical Physics , vol. 347, no. 1, pp. 291– 313, 2016

    2016

  15. [23]

    M. M. Wilde, Quantum Information Theory. Cambridge: Cambridge University Press, 2013

    2013

  16. [24]

    Coding theorem and strong converse for quantum chan- nels,

    A. Winter, “Coding theorem and strong converse for quantum chan- nels,” IEEE Trans. Inf. Theory , vol. 45, no. 7, pp. 2481–2485, 1999. [Online]. Available: http://ieeexplore.ieee.org/document/796385/

    1999

  17. [25]

    Compound channel capacities under energy constraints and application,

    A. Cacioppo, J. N ¨otzel, and M. Rosati, “Compound channel capacities under energy constraints and application,” in 2021 IEEE International Symposium on Information Theory (ISIT) , 2021, pp. 640–645

    2021

  18. [26]

    Arbitrarily varying and compound classical-quantum channels and a note on quantum zero-error capacities,

    I. Bjelakovic, H. Boche, G. J. en, and N. J., “Arbitrarily varying and compound classical-quantum channels and a note on quantum zero-error capacities,” Aydinian H., Cicalese F ., Deppe C. (eds) Information Theory, Combinatorics, and Search Theory. Lecture Notes in Computer Sc...

    2013

  19. [27]

    General properties of entropy,

    A. Wehrl, “General properties of entropy,” Rev. Mod. Phys., vol. 50, pp. 221–260, Apr 1978. [Online]. Available: https://link.aps.org/doi/10.1103/RevModPhys.50.221

    1978 doi

  20. [28]

    Infinite-fold quantum advantage in classical correlation sensing,

    J. N ¨otzel and P. Munar-Vallespir, “Infinite-fold quantum advantage in classical correlation sensing,” 2025. [Online]. Available: https://arxiv.org/abs/2503.17235

    2025

  21. [29]

    Csisz ´ar and J

    I. Csisz ´ar and J. K ¨orner, Information Theory: Coding Theorems for Discrete Memoryless Systems , 2nd ed. Cambridge University Press, 2011. APPENDIX We now provide the proofs of the lemmas. Proof of Lemma 6: It holds 1 = Q + Q⊥ for some orthogonal projection Q⊥. Let Q(i) := ...

    2011

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.