{"id":"d501b0d4-e64e-4f83-a85e-04fa4efb0391","arxiv_id":"2411.11604","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The joint communication and sensing region of the lossy bosonic channel is the rectangle [0,g(νE)] × [0,(E/2) min |k−k'|²], showing no tradeoff between the two tasks.","lead":"This paper asks how well one laser pulse can simultaneously carry data to a receiver and measure how much of the pulse reflects back to the sender. For an idealized lossless model, the answer is a rectangle: both tasks hit their individual quantum limits at once, with a fourfold quantum advantage in sensing accuracy over homodyne detection.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The region formula uses ν as an intensity in g(νE), but Definition 1 defines the forward output as |να⟩, making ν an amplitude; for the channel as written the correct capacity is g(|ν|²E), so the stated communication rate is off by a square.","rationale":"I read the paper as establishing that, for the proposed bidirectional lossy bosonic channel, there is no communication-sensing tradeoff: the achievable region is a rectangle bounded by the Holevo capacity g(νE) and the quantum Chernoff exponent E/2 min_{k≠k'}|k−k'|². The reader's verdict is CONDITIONAL, citing two issues: the channel model uses ν inconsistently (amplitude vs intensity) and never states the physical relation between k and ν, and the random-coding construction does not guarantee every codeword saturates the energy bound required for Dmax. I agree that the ν inconsistency is real and load-bearing, because it directly affects the headline formula in Theorem 2 and the classical comparison in Section IV. The reader's second concern about constant-energy codewords is less severe: in a random Gaussian codebook with M≈2^{nR} codewords, the minimum per-symbol energy among codewords is F−O(√(R/n)) by standard concentration of chi-squared variables, so every codeword asymptotically reaches E−ε; choosing ε→0 recovers Dmax. Thus that gap is fixable by a standard argument and does not threaten the no-tradeoff structure. The most serious issue is the amplitude/intensity mismatch: for the channel defined with |να⟩, the forward capacity is g(|ν|²E), not g(νE). This is an internal inconsistency, not merely a disagreement with outside consensus, and it changes the numerical content of the theorem unless the parametrization is corrected. Because the correction is straightforward and the rectangular tradeoff-free structure likely survives the reparametrization, I do not change the reader's CONDITIONAL verdict; I would keep it conditional pending the fix and an explicit statement of the physical constraint between k and ν (e.g., k²+|ν|²≤1 for a passive lossy splitter).","tokens_in":10042,"tokens_out":7665,"duration_ms":76524,"concrete_test":"Verify by direct substitution: for the channel W_{ν,k}(α)=|kα⟩⊗|να⟩ in Definition 1, compute the forward output photon number ⟨να|a†a|να⟩=|ν|²|α|². Compare with the known pure-loss capacity formula C=g(ηE), where η is intensity transmissivity. The theorem's rate g(νE) matches the definition only if η=ν, i.e., if the output amplitude were √ν α. Re-derive Theorem 2 and the Section IV classical comparison with the corrected forward transmissivity |ν|² and SNR=|ν|²E; if the rates change, the manuscript must fix the parameter convention before the numerical claims can be assessed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Definition 1 states W_{ν,k}(α)=|kα⟩⊗|να⟩. In the coherent-state notation, the mean photon number of the forward output is |ν|²|α|², so ν is an amplitude scaling and the intensity transmissivity of the forward path is |ν|². The known capacity of a pure-loss bosonic channel with intensity transmissivity η and input energy E is g(ηE). Therefore Theorem 2's communication bound g(νE) does not match the channel defined in Definition 1; it would be correct only if the forward output were |√ν α⟩ or if ν were explicitly defined as the square root of intensity transmissivity. The same ambiguity propagates into Section IV: the classical comparison uses SNR=νE, but homodyne detection on |να⟩ yields a signal amplitude ν Re(α), giving SNR=|ν|²E. This is not a harmless convention choice: g(νE)≠g(|ν|²E) for 0<ν<1, so the quantitative region and the claimed factor-of-4 advantage are not established for the channel as defined. The detection exponent Dmax is unaffected because it depends only on k and E, but the communication axis of the central rectangle is wrong unless the parametrization is fixed. A one-line correction (replace ν by √ν in Definition 1, or replace g(νE) by g(|ν|²E) everywhere) would resolve it, but as written the theorem and the comparison contain a genuine inconsistency.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript introduces a bidirectional lossy bosonic quantum channel (BLBC) in which coherent-state symbols are sent to a receiver while a backreflected coherent field with reflectivity k is used for sensing. The main result (Theorem 2) claims that the achievable communication–detection region is rectangular: any rate up to the Holevo capacity g(νE) and any detection exponent up to Dmax = E min_{k≠k'} |k−k'|²/2 are simultaneously achievable, with no tradeoff. The proof combines known results on the pure-loss channel capacity, a finite-dimensional truncation argument (Lemmas 1–3), the multiple quantum Chernoff bound of [14], and a converse adapted from [7]. A comparison with a classical AWGN/homodyne model is used to claim a fixed factor-of-4 detection advantage and an unbounded communication advantage.","tokens_in":10283,"tokens_out":37541,"duration_ms":364531,"significance":"If the technical issues are resolved, this would be a useful first continuous-variable result on joint communication and sensing, showing that for this coherent-state model the two tasks do not compete. Strengths of the paper include the assembly of the central region from external theorems with no fitted parameters, an explicit truncation strategy, a concrete falsifiable detection-exponent formula, and an honest statement of limitations (no thermal noise, non-constructive POVM). However, several load-bearing points in the current version prevent the claims from being accepted as stated.","major_comments":[{"comment":"The parameter ν is used as an amplitude in Definition 1 (forward output |να⟩) but as an intensity transmissivity in the capacity expression g(νE). With the channel as defined, the forward intensity transmissivity is |ν|², so the correct communication bound is g(|ν|²E), not g(νE). The same inconsistency enters Section IV, where the classical SNR is quoted as νE although homodyne detection on |να⟩ gives a signal amplitude ν Re(α) and hence SNR |ν|²E. This is load-bearing because Theorem 2 and the quantitative comparison depend on it; please either replace the forward output by |√ν α⟩ or use g(|ν|²E) throughout.","section":"Definition 1, Theorem 2, Section II.A"},{"comment":"The coherent-state Chernoff exponent is quoted as |k'α−kα|²/2. With the paper's own definition D(ρ,σ)=sup_s −log Tr(ρ^s σ^{1−s}), the exact value for two pure coherent states is obtained from |⟨kα|k'α⟩|² = exp(−|k−k'|²|α|²), so D = |k−k'|²|α|², not half of that. Consequently Dmax should be E min_{k≠k'}|k−k'|², a factor of 2 larger than stated. This also changes the claimed quantum advantage over the classical D=E/8 from a factor of 4 to a factor of 8 for K={0,1}. Please correct Eq. (18) (or the conversion from the quadrature-space formula in [23]) and propagate the correction through Theorem 1, Theorem 2, and the converse.","section":"Section V.C.1, Eq. (18), Theorem 2"},{"comment":"Detection error in Definition 3 is worst-case over messages. The proof obtains the exponent by replacing Σ_x N(α_x|α^n)|α_x|² with its expectation E via the law of large numbers. For an i.i.d. random codebook of size exp(nR), the minimum-energy codeword has per-symbol energy E−δ_R for a positive δ_R (large-deviations lower tail), so the worst-case exponent is strictly below Dmax. The sentence 'the performance of the code is given by the codeword with the least amount of energy' recognizes this, but the proof does not show that a code with all codewords at exactly energy E also achieves the communication rate g(νE). Please provide an explicit constant-composition/constant-energy code construction and prove that its communication rate tends to g(νE).","section":"Section V.C.1, Eqs. (19)-(20), Definition 3"},{"comment":"The classical comparison model is internally inconsistent. Definition 5 defines Q_{ν,k} with outputs Y_Ai=x_i+Z_Ai and Y_Bi=k x_i+Z_Bi, and the text calls Y_A the reflected part; but for the BLBC under homodyne detection with ν=1, the reflected output has amplitude k x_i, not x_i, so the labels are interchanged. The region stated after Theorem 3, R≤1/2 log(1+νE) and D≤E/8, does not follow from Definition 5 as written (which would give R≤1/2 log(1+E) and D≤E/8 for k=1, with no ν dependence). The model and parameter mapping need to be restated before the claimed factor-of-4 comparison can be assessed.","section":"Definition 5, text after Theorem 3"}],"minor_comments":[{"comment":"The definition of the Gordon function has a sign typo: g(x) is written with '+ x log x' but Theorem 2 uses the correct '− x log x'.","section":"Section II.A"},{"comment":"The indices in the decoding and detection POVMs are inconsistent: the text defines {Λ_k}_{j=1}^M and {Π_k}_{k∈K}, while Eqs. (1)-(2) use Λ_m and Π_{k,m}. Please clarify that the detection POVM may depend on the message m and fix the notation.","section":"Definition 3"},{"comment":"The physical relation between the amplitude coefficients k and ν is never stated. For a passive beam splitter, k²+ν²≤1; if this is intended, it should be stated, since Theorem 2 and the comparison treat them as independent parameters.","section":"Definitions 1-2"},{"comment":"The remark on approximating continuous measures by finite discrete measures is very terse; please spell out why the setwise convergence preserves the Holevo information within ε and how the finite alphabet size is chosen.","section":"Section V.A, Remark"},{"comment":"The converse is only a paragraph; please expand the adaptation of [7] to infinite-dimensional coherent-state channels, in particular the handling of the energy constraint and the uniform prior over K.","section":"Section V.C.2"}],"recommendation":"major_revision","confidential_remarks":"I want to flag for the editor that the two quantitative errors (the ν amplitude/intensity mismatch and the factor-of-2 in the Chernoff exponent) appear to be convention/notation errors rather than deliberate modeling choices, but they do change the main quantitative claims. The worst-case-energy gap in the achievability proof is more than cosmetic and should be fixed with an explicit constant-energy code argument. I would not reject the paper on these grounds, but it should not be published until the region formula and the classical comparison are corrected."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my take. The paper is close to a solid contribution, but the exact statement of the main theorem has a parametrization bug that has to be fixed. What's genuinely new: a closed-form achievable region for a lossy, noiseless bidirectional bosonic channel, extending the finite-dimensional result of Wang et al. to infinite dimensions. The region is rectangular: communication rate up to g(νE), sensing exponent up to (E/2) min|k−k'|², and no tradeoff between the two. That is a clean benchmark for quantum JCAS, and the proof is mostly assembled from known ingredients: Holevo capacity for the forward link, the multiple Chernoff bound for sensing, and a finite-dimensional truncation to make the machinery apply. The paper does a good job of showing the same constant-energy codebook can achieve both extremes, at least in spirit.\n\nThe soft spots are manageable. First, Definition 1 defines the forward output as |να>, so ν is an amplitude; but the rest of the paper treats ν as an intensity transmissivity in g(νE) and in the classical SNR comparison. For the channel as written, the communication rate should be g(|ν|²E), not g(νE). This is a genuine inconsistency, not a harmless convention choice, because it changes the numbers in Theorem 2 and Figure 1. The factor-4 detection advantage does not depend on ν, and the qualitative unbounded communication advantage survives the correction, so the main conclusions are robust.\n\nSecond, the detection achievability proof samples a random codebook with average energy E and then notes that the worst-case codeword determines the sensing exponent. That only gives a bound based on the minimum codeword energy, not E; the claimed D = E/2 requires every codeword to have energy close to E. The paper mentions constant-energy codes but does not actually construct one in the proof. A spherical or constant-composition codebook fixes this in a few lines, because uniform measure on the energy sphere has the same asymptotic marginal as the Gaussian ensemble, so the Holevo rate is preserved. Minor, but it needs to be written out.\n\nThird, the detection converse is waved through by reference to [7] without an explicit infinite-dimensional argument. The claim is plausible, but it deserves a few sentences. The citation pattern is otherwise fine: core results are properly attributed to external theorems, and the self-citations support only auxiliary statements.\n\nWho is this for: people working on quantum JCAS, optical communication limits, and c-q capacity regions in infinite dimensions. It is a useful example rather than a breakthrough. I'd send it to peer review; the flaws are fixable and the benchmark is worth having on record. I'd also bring it to a reading group to discuss the codebook construction issue.","headline":"Useful closed-form JCAS region for a bosonic channel, but the ν vs ν² parametrization and the missing constant-energy codebook proof need fixing before the exact statement can be trusted.","tokens_in":10869,"tokens_out":7924,"would_cite":true,"duration_ms":84863,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P45","94A40"],"pacs":[],"model":"deepseek-v4-flash","headline":"For a noiseless bosonic channel with backreflection, communication and sensing need not compete: the same code achieves the Holevo capacity and the optimal Chernoff detection exponent.","keywords":["joint communication and sensing","lossy bosonic channel","coherent states","quantum Chernoff bound","Holevo capacity","quantum advantage","reflectivity estimation","classical-quantum channel"],"falsifier":"Measure the reflected field of a real beam splitter fed by a coherent state: if the reflected state contains thermal noise above shot noise, or if its strength is tied to the forward strength by a passive-splitter relation such as $k=\\sqrt{1-\\nu^2}$, then the predicted region $g(\\nu E)\\times \\frac{E}{2}\\min_{k\\ne k'}|k-k'|^2$ will not be observed. A direct laboratory test would prepare $K=\\{0,1\\}$ with known energy $E$ and use collective measurements on the backreflected beam to see whether the error exponent reaches $E/2$ or stops at the homodyne value $E/8$.","tokens_in":9780,"feed_emoji":"📡","tokens_out":15198,"duration_ms":136758,"temperature":0.7,"pith_summary":"This paper studies a noiseless bosonic channel in which an input coherent state $|\\alpha\\rangle$ splits into a forward beam $|\\nu\\alpha\\rangle$ sent to a receiver and a backreflected beam $|k\\alpha\\rangle$ returned to the sender, with $k$ an unknown reflectivity the sender wants to estimate while also sending data. The authors prove that for any finite family of reflectivities $k\\in K$, the achievable pairs (communication rate $R$, detection exponent $D$) form exactly the rectangle $0\\le R\\le g(\\nu E)$ and $0\\le D\\le \\frac{E}{2}\\min_{k\\ne k'}|k-k'|^2$, where $E$ is the average input energy. In words, there is no tradeoff: a single codebook with constant per-codeword energy simultaneously reaches the Holevo capacity (the best possible rate for sending classical data over a quantum channel) and the quantum Chernoff bound (the optimal error exponent for distinguishing quantum states). The paper also shows that optimal quantum measurements beat shot-noise-limited homodyne detection by a fixed factor of four in detection exponent, while the communication advantage grows without bound as the received photon number goes to zero. This matters because it identifies the pure-loss bosonic channel as a clean case where joint communication and sensing costs nothing in rate or sensing quality.","feed_headline":"No tradeoff: one code senses and communicates at quantum limits","feed_subtitle":"Quantum-optimal measurements hit Holevo capacity and Chernoff sensing bound at once, beating homodyne by 4x.","key_machinery":"The load-bearing object is the bidirectional lossy bosonic channel (BLBC), a classical–quantum channel $W_{\\nu,k}(\\alpha)=|k\\alpha\\rangle\\otimes|\\nu\\alpha\\rangle$ that couples one coherent input to an estimated backreflection and a forward data beam. The proof machinery combines the Holevo-capacity formula $g(\\nu E)$ for the pure-loss bosonic channel; the quantum Chernoff bound for binary coherent-state discrimination, extended to many hypotheses through the multiple Chernoff distance; a finite-dimensional approximation lemma that projects long sequences of coherent states onto $\\lfloor\\log n\\rfloor$ Fock levels, making the finite-dimensional JCAS achievability proof applicable; and continuity of the Chernoff exponent under trace-norm convergence to lift the result back to the infinite-dimensional setting. These pieces fit together so that the same random codebook and the same measurement sequence attain both extremes simultaneously.","core_discovery":"The central claim is Theorem 2: for a $K$-family of bidirectional lossy bosonic channels, defined by the map $\\alpha\\mapsto |k\\alpha\\rangle\\otimes|\\nu\\alpha\\rangle$, the achievable communication–detection region is the full rectangle $[0,g(\\nu E)]\\times[0,D_{\\max}]$ with $D_{\\max}=\\frac{E}{2}\\min_{k\\ne k'}|k-k'|^2$. The forward communication task is a classical–quantum channel whose capacity is the Holevo quantity, equal to $g(\\nu E)$ for a lossy bosonic channel with transmissivity $\\nu$ and mean input energy $E$; the sensing task is $|K|$-ary discrimination of the backreflected coherent states $|k\\alpha\\rangle$, whose optimal error exponent is a multiple quantum Chernoff distance. The region is proved in both directions: achievability by random coding with a Gaussian-like input distribution and a finite-dimensional approximation that projects onto the lowest $\\lfloor\\log n\\rfloor$ Fock levels, importing the multiple-Chernoff bound for general quantum measurements (POVMs), and the converse by noting that the earlier finite-dimensional JCAS argument applies to coherent states. The result is that energy, not any tradeoff, is the single resource limiting both tasks, and optimal codes are constant-energy codes.","pith_inferences":["If the BLBC is realized by a passive beam splitter, unitarity forces $k^2+\\nu^2\\le 1$ (or one of the two parameters must be read as an intensity transmissivity), whereas the paper leaves $k$ and $\\nu$ independent and switches between amplitude and intensity readings of $\\nu$; re-deriving the region under the passive-splitter constraint is a direct test of how generic the rectangle is.","The no-tradeoff result likely depends on the coherent-state, noiseless structure; adding thermal noise to the backreflected mode should produce a genuine rate–exponent tradeoff, because the Chernoff exponent of displaced thermal states depends on temperature and the forward capacity decreases.","A natural extension is to replace the finite set $K$ by a continuum of reflectivities and use Bayesian or local estimation instead of Chernoff exponents; the expected behavior is that the rectangular region turns into a curve governed by the energy constraint.","Because the sensing POVM is non-constructive, a practically testable question is how closely heterodyne or other feasible receivers approach $E/2$ for finite blocklengths; the homodyne comparison already sets a factor-of-four gap to close."],"forward_implications":["For a two-reflectivity setup ($K=\\{0,1\\}$), the sensing exponent is $E/2$ with optimal quantum measurements versus $E/8$ with homodyne detection: a fixed factor of four that does not depend on photon number.","The communication rate saturates the Holevo capacity $g(\\nu E)$, which exceeds the homodyne/Shannon rate $\\frac{1}{2}\\log(1+\\nu E)$ and becomes unboundedly larger as $E\\to 0$, so low-photon regimes are where quantum JCAS pays off most.","No time-sharing between communication-optimal and sensing-optimal codes is needed; there is a single family of constant-energy codes that are optimal for both tasks at once.","Any attempt to exceed either $g(\\nu E)$ or $D_{\\max}$ fails: because the region is a rectangle, the two tasks do not compete for resources beyond the shared energy budget.","The optimal receiver POVM is not constructed and may be arbitrarily complex, so practical implementations will initially fall short of the predicted region until suitable collective measurements are realized."],"supporting_citations":[{"why":"Supplies the finite-dimensional joint quantum communication and sensing framework whose achievability and converse arguments the paper adapts to the bosonic channel.","marker":"[7]"},{"why":"Provides the ultimate classical communication rate $g(\\nu E)$ of a lossy bosonic channel, setting the communication side of the region.","marker":"[11]"},{"why":"Gives the multiple quantum Chernoff distance and the POVM achievability bound used to bound the sensing error exponent for $|K|$ reflectivities.","marker":"[14]"},{"why":"Defines the classical AWGN joint communication and detection model used as the homodyne benchmark, giving the region $R\\le\\frac{1}{2}\\log(1+\\nu E)$, $D\\le E/8$.","marker":"[15]"},{"why":"Establishes the unbounded quantum communication advantage at low received photon numbers that the comparison with the classical region relies on.","marker":"[18]"},{"why":"Provides the closed-form Chernoff exponent for displaced Gaussian states, from which the coherent-state exponent $|d|^2/2$ is read off.","marker":"[23]"},{"why":"Shows achievability of the quantum Chernoff bound for discriminating two hypotheses, the base case for the sensing exponent.","marker":"[12]"},{"why":"Proves optimality of the quantum Chernoff bound for symmetric binary hypothesis testing, supporting the converse for coherent states.","marker":"[13]"}],"fun_headline_variants":["No tradeoff: one code achieves both quantum limits","Joint comms and sensing: quantum limits met together","One energy budget, two quantum tasks, no compromise","Quantum-optimal messages and measurements in one code","Bosonic channel: simultaneous Holevo and Chernoff bounds"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result stands or falls with the assumption that the reflected and forward beams are exactly two independent ideal laser-like quantum states, with no added noise and no required relation between their strengths; if a real reflector is noisy or if the two strengths are forced to obey a passive-splitter constraint, the rectangle and the fourfold advantage can collapse.","fun_headline_variants_meta":{"raw":{"variants":["No tradeoff: one code achieves both quantum limits","Joint comms and sensing: quantum limits met together","One energy budget, two quantum tasks, no compromise","Quantum-optimal messages and measurements in one code","Bosonic channel: simultaneous Holevo and Chernoff bounds"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000186,"raw_usage":{"total_tokens":1318,"prompt_tokens":933,"completion_tokens":385,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":549,"completion_tokens_details":{"reasoning_tokens":308}},"tokens_in":549,"tokens_out":385,"duration_ms":50468,"temperature":1.0,"reasoning_tokens":308,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T18:21:46.177839+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the reflected field of a real beam splitter fed by a coherent state: if the reflected state contains thermal noise above shot noise, or if its strength is tied to the forward strength by a passive-splitter relation such as $k=\\sqrt{1-\\nu^2}$, then the predicted region $g(\\nu E)\\times \\frac{E}{2}\\min_{k\\ne k'}|k-k'|^2$ will not be observed. A direct laboratory test would prepare $K=\\{0,1\\}$ with known energy $E$ and use collective measurements on the backreflected beam to see whether the error exponent reaches $E/2$ or stops at the homodyne value $E/8$.","supporting_citations":[{"cited_title":"Ultimate classical communication rates of quantum optical channels,","cited_arxiv_id":null,"evidence_quote":"Provides the ultimate classical communication rate $g(\\nu E)$ of a lossy bosonic channel, setting the communication side of the region."},{"cited_title":"Discriminating quantum states: The multiple chernoff distance,","cited_arxiv_id":null,"evidence_quote":"Gives the multiple quantum Chernoff distance and the POVM achievability bound used to bound the sensing error exponent for $|K|$ reflectivities."},{"cited_title":"Joint communication and binary state detection,","cited_arxiv_id":null,"evidence_quote":"Defines the classical AWGN joint communication and detection model used as the homodyne benchmark, giving the region $R\\le\\frac{1}{2}\\log(1+\\nu E)$, $D\\le E/8$."},{"cited_title":"Operating Fiber Networks in the Quantum Limit","cited_arxiv_id":"2201.12397","evidence_quote":"Establishes the unbounded quantum communication advantage at low received photon numbers that the comparison with the classical region relies on."},{"cited_title":"Quantum chernoff bound as a measure of distinguishability between density matrices: Application to qubit and gaussian states,","cited_arxiv_id":null,"evidence_quote":"Provides the closed-form Chernoff exponent for displaced Gaussian states, from which the coherent-state exponent $|d|^2/2$ is read off."}],"review_version":1}