{"id":"df0d60a6-cb99-485d-9a59-8d7fd5f57ce2","arxiv_id":"1908.03058","paper_version":3,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Microwave quantum illumination with a Josephson parametric converter and a digital phase-conjugate receiver detects a room-temperature target at 1 meter and, under simulated perfect idler photon counting, shows up to 4 dB advantage over classical benchmarks.","lead":"Scientists sent entangled microwave photon pairs through a one-meter room-temperature radar link and used a digital receiver to show that the quantum-correlated probe beats a classical noise radar, with the strongest advantage appearing only when a perfect idler detector is simulated rather than actually built. The result is a proof of principle that microwave quantum illumination, a proposed low-power sensing technique, can work outside the cryostat in a noisy environment.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The calibrated 1-4 dB quantum advantage hinges on a linear idler-noise subtraction in Eq. 22; the paper does not show this equals an ideal photon-counting receiver at the covariance level.","rationale":"The reader's weakest_assumption is exactly the idler calibration; I agree. This is the load-bearing point because the paper's strongest claim—quantum advantage over coherent homodyne, the optimal classical benchmark—is presented only under the calibrated dashed curves. Those curves are inferred, not measured, and the inference step is a single variance subtraction described in one sentence. Eq. (22)'s moment formula depends on the whole Gaussian covariance matrix; using the noisy idler mode for some moments and the calibrated idler for one term can be inconsistent. A covariance-level test can settle this. If that test reproduces the dashed curves, the extrapolation is sound and ACCEPT is justified. If not, the claim should be conditional on reworking the calibration or weakening the conclusion. I would not reject: the raw QI/CI comparison and the entanglement witness are measured, the limitation is openly stated, and the authors creditably do not claim the uncalibrated digital receiver beats homodyne. But the advertised advantage is currently a model-dependent extrapolation, so a conditional verdict is the most honest adjustment.","tokens_in":16011,"tokens_out":16795,"duration_ms":192592,"concrete_test":"Recompute the dashed calibrated QI curves in Fig. 2(b) from the measured 4x4 covariance matrix after applying the inverse idler-amplifier map consistently: scale the idler variances and cross-correlations by 1/G_I and subtract (nadd,I + 1) from the idler variance before evaluating the Gaussian photon-counting SNR. Compare these values with the published Eq. (22) results; if any point differs by more than the experimental 95% confidence intervals, the calibrated 1-4 dB advantage is not equivalent to an ideal idler photon-number receiver.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The direct raw QI-vs-CI comparison (Fig. 2b, solid lines) is measured and robust, but the headline quantum advantage over coherent-state homodyne (about 1 dB) and heterodyne (about 4 dB) is not measured. It is obtained by replacing the measured idler photon number with the calibrated value ⟨a†_I a_I⟩ = ⟨a†_{det,I} a_{det,I}⟩/G_I − (nadd,I + 1) inside the receiver SNR formula, Methods V Eq. (22). The SNR of the phase-conjugate receiver, Eq. (19), is a function of fourth-order photon-number moments of the modes after the beam splitter, not of the idler mean alone. For the Gaussian states used here those moments are determined by the full signal-idler covariance matrix; a phase-preserving idler amplifier changes the idler marginal variance and the signal-idler correlations in a coupled way. Subtracting (nadd,I + 1) from one idler moment reproduces a perfect idler photon counter only if the entire measured covariance matrix is first transformed by the inverse amplifier model. The manuscript does not display that covariance-level transformation, and it explicitly concedes that the digital receiver cannot beat homodyne without this calibration. Thus the central quantum-advantage claim rests on an unverified linear subtraction, not on a direct photon-counting measurement.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a microwave quantum illumination (QI) experiment in which a Josephson parametric converter generates entangled signal-idler fields, the signal probes a room-temperature target (including a free-space link at 1 m), and a digital phase-conjugate receiver is implemented from linear quadrature measurements. The central experimental results are a measured SNR comparison between QI and a symmetric classically-correlated illumination (CI) benchmark, showing up to 3 dB raw advantage of QI at low photon numbers, and an inferred/simulated SNR for perfect idler photon-number detection, which the authors report as up to 4 dB over coherent heterodyne and about 1 dB over coherent homodyne detection. The paper explicitly states that without the idler calibration the digital receiver cannot outperform coherent homodyne, and that in the implemented setup passive detection of the amplifier noise would give a much higher SNR than any of the compared active protocols.","tokens_in":16321,"tokens_out":8664,"duration_ms":97392,"significance":"If the results hold, this is a valuable experimental milestone: it is a microwave QI demonstration with a room-temperature target, independent calibration of gains and added noise, direct measurement of the QI-versus-CI SNR difference, and a careful separation of measured raw performance from simulated ideal-idler performance. The entanglement verification and the explicit acknowledgment of the passive-detection limitation are commendable. The main significance beyond the raw comparison is the claimed simulated quantum advantage over coherent-state benchmarks, and that claim is the part that needs the closest scrutiny.","major_comments":[{"comment":"The calibrated/ideal-idler SNR is obtained by replacing the idler photon number in Eq. (22) with <a†_I a_I> = <a†_{det,I} a_{det,I}>/G_I − (nadd,I + 1), while the other moments in Eq. (22), namely <N_i,+> and <N_i,−>, are evidently taken from the measured detected modes. For a phase-preserving idler amplifier, the transformation from the JPC-output idler mode to the detected idler mode also rescales signal-idler cross-correlations and modifies the fourth-order moments that enter <N_i,+> and <N_i,−>. Subtracting a single mean value from one term of Eq. (22) is not equivalent to inverse-calibrating the full two-mode covariance matrix. The manuscript should either provide the covariance-level inverse transformation and evaluate Eq. (22) entirely in terms of the calibrated JPC-output parameters, or explicitly relabel the dashed curves in Figs. 2(b) and 3 as a model-dependent extrapolation rather than an inferred measured SNR. As written, the claimed 1 dB advantage over homodyne and 4 dB advantage over heterodyne are not supported by the displayed calculation.","section":"Methods V, Eq. (22); main text around Fig. 2(b)"},{"comment":"The calibrated QI advantage over coherent homodyne is reported as approximately 1 dB in the region N_S < 0.4, but the manuscript does not provide an uncertainty estimate on this difference or on the inferred SNR values used to establish it. Since Fig. 3(b) omits the homodyne benchmark because the expected advantage is said to be smaller than systematic errors, the reader cannot tell whether the 1 dB advantage in Fig. 2(b) is statistically significant. The authors should report the inferred SNR difference with propagated statistical and systematic uncertainties, or temper the claim accordingly.","section":"Fig. 2(b) and text near 'approximately 1 dB larger'"}],"minor_comments":[{"comment":"The calibration formula is written as <a†_I a_I> = <a†_{det,i} a_{det,i}>/G_I − (nadd,I + 1); the index i is not defined there and the subscript on the detected mode does not match the left-hand side. Please clarify the notation and reconcile the '+1' with the definition of nadd,I in the Supplementary Information, Eq. (11).","section":"Main text, paragraph after Eq. (3)"},{"comment":"The legend uses 'raw SNR' and 'calibrated idler' but the caption does not define these terms; please state explicitly that 'raw' means no idler calibration and 'calibrated idler' means the simulated perfect-idler-photon-detection result.","section":"Fig. 2(b) caption and legend"},{"comment":"The figure shows both object distance and total loss on the horizontal axis, but the caption does not explain how the loss values are derived from the free-space distances; please add a sentence describing the loss model used.","section":"Fig. 3(b)"},{"comment":"The abstract describes the advantage as being 'compared to the relative classical benchmark'; since the only measured advantage is against symmetric CI and the coherent-state advantages are simulated, please state this distinction explicitly in the abstract as well as in the conclusion.","section":"Abstract and Conclusion"},{"comment":"Equation (3) in the main text and Eq. (19) in the Methods are the same expression; please use a single equation number or cross-reference to avoid duplication.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The raw QI-versus-CI comparison is a solid experimental result and is appropriately calibrated, but the headline simulated quantum advantage over coherent-state benchmarks depends on an idler-noise subtraction whose covariance-level validity is not established. This is fixable by providing a full derivation or by reframing the claims, so major revision is more appropriate than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Natasha,\n\nShort version: this is a worthwhile experimental paper, and you should send it to referees if you handle it, but the headline quantum advantage over coherent-state homodyne is an inference, not a measurement, and the calibration step deserves a careful look.\n\nWhat's actually new: the authors implement a quantum illumination link at microwave frequencies using a Josephson parametric converter source, a one-meter free-space room-temperature target, and a digital phase-conjugate receiver built in post-processing. The raw QI-vs-symmetric-CI comparison is directly measured, with the same apparatus and calibrated gains, and shows up to 3 dB advantage at low photon numbers. That part is solid. They also explicitly label the stronger claims—4 dB over heterodyne, 1 dB over homodyne—as simulations assuming perfect idler photon-number detection, obtained by calibrating the idler amplifier noise. That transparency matters.\n\nThe soft spots are real but proportionate. The main one is that the calibrated advantage is not a direct measurement. The paper replaces the measured idler photon number with a calibrated noiseless value inside the SNR variance formula, and it does not show the full covariance-level transformation. The stress-test note worries this is unjustified. I think that concern is partly answered by the structure of the problem: the idler is one half of a two-mode squeezed vacuum, so its reduced state is thermal, and for a thermal state the photon-number statistics are fixed by the mean. The amplifier adds thermal noise, so subtracting the added mean noise and using the calibrated cross-correlation (which they do) may be sufficient. But the paper would be stronger if it displayed the covariance inversion explicitly and justified why the linear subtraction reproduces an ideal photon counter. As it stands, the 1-4 dB advantage over coherent benchmarks is an extrapolation, not an observation.\n\nOther minor issues: no data or code artifacts are provided, so the experiment cannot be independently re-analyzed from the manuscript. The theory curves share parameters with the measured data, which is fine for consistency checks but not an independent verification. The authors are honest that, in the current setup, a passive detection of the amplifier noise would be a stronger classical strategy; that limits the practical significance but not the fundamental demonstration.\n\nWho is this for: anyone working in quantum sensing, particularly microwave QI and continuous-variable quantum information. It is a clear step beyond previous microwave QI demonstrations (refs. 20-21) in the direction of a room-temperature free-space target and a digital receiver.\n\nRecommendation: accept for peer review. The raw comparison is measured, the calibration is described in enough detail to be challenged, and the limitations are stated. A referee should push for a clearer derivation of the calibrated SNR and ideally for a direct test of the calibration against a known state. But this is not a desk-reject.","headline":"A solid experimental step for microwave quantum illumination: the raw QI-vs-CI advantage is measured, but the stronger advantage over coherent homodyne is a clearly labeled simulation that depends on the idler calibration.","tokens_in":16819,"tokens_out":9673,"would_cite":true,"duration_ms":101417,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Microwave quantum illumination beats classical noise radar","keywords":["quantum illumination","microwave radar","Josephson parametric converter","phase-conjugate receiver","entanglement","target detection","quadrature measurement","noise radar"],"falsifier":"Replace the calibrated idler post-processing with a real microwave photon-number-resolving detector at the JPC output and measure the SNR for the same target at $N_S<0.4$ photons per mode; if the realized advantage over coherent-state homodyne detection does not reach the predicted ~1 dB (or the heterodyne advantage does not reach ~4 dB), the simulated idler calibration is the point of failure.","tokens_in":15830,"feed_emoji":"📡","tokens_out":9477,"duration_ms":94101,"temperature":0.7,"pith_summary":"Microwave quantum illumination, a sensing protocol that uses entangled photon pairs to detect faint objects against bright thermal noise, is brought into the microwave domain with a Josephson parametric converter source and a fully digital receiver. The paper reports that this receiver outperforms a symmetric classically correlated noise radar by up to 3 dB in raw signal-to-noise ratio at low photon numbers, even though the entanglement is destroyed on the round trip to a room-temperature object one meter away. Because a digital linear receiver captures the idler's vacuum noise, the raw receiver cannot beat coherent-state homodyne detection. When the measured idler statistics are calibrated to simulate a perfect photon-number-resolving idler detector, the inferred advantage grows to about 1 dB over homodyne and up to 4 dB over heterodyne detection. This suggests that low-power, non-invasive microwave sensing could eventually be a practical application of quantum correlations.","feed_headline":"Microwave quantum illumination beats classical noise radar","feed_subtitle":"A Josephson source plus digital receiver yields 3 dB raw, up to 4 dB with ideal idler counting.","key_machinery":"The central object is the Josephson parametric converter (JPC), a superconducting three-wave-mixing amplifier that emits two-mode squeezed microwave fields with a non-zero phase-sensitive cross-correlation between signal and idler; it supplies the entangled source. The carrying mechanism is the digital phase-conjugate receiver: the returned signal is phase-conjugated, mixed with the retained idler on a 50-50 beam splitter, and the photon-count difference is used as the test statistic, with all operations performed in post-processing on digitized quadrature measurements. The SNR formula for this receiver is evaluated from the measured covariance matrix, and the simulated ideal-idler result follows from reducing the measured idler variance by the calibrated added noise and vacuum contribution. The receiver's job is to convert the surviving signal-idler correlations into a detection advantage after entanglement has been broken by loss and noise.","core_discovery":"The paper's central claim is that quantum illumination works at microwave frequencies under realistic conditions: a Josephson parametric converter produces entangled signal and idler fields at 10.09 and 6.8 GHz, the signal is amplified and sent through a free-space link to a room-temperature copper target, and a digital phase-conjugate receiver reconstructs the full covariance matrix from linear quadrature measurements. Using the resulting SNR for the binary target-absence/presence decision, the authors find that quantum illumination beats a symmetric classically correlated noise radar by up to 3 dB at low signal photon numbers, and that the measured data are consistent with theory down to less than one photon per mode. They then extrapolate from the measured idler noise to the case of an ideal idler photon-number detector and report an advantage over coherent-state heterodyne detection of up to 4 dB and over coherent-state homodyne detection of about 1 dB in the entangled regime. The paper is explicit that without this calibration the raw digital receiver cannot outperform coherent homodyne detection, so the quantum advantage is demonstrated only under the assumption of perfect idler photon counting.","pith_inferences":["The paper's own calibration step is the load-bearing extrapolation: a direct test would build a real microwave photon-number-resolving detector on the idler arm and compare the achieved SNR with the calibrated curve; if real detector inefficiency or dark counts degrade the result, the 1-4 dB advantage may not survive.","Because the demonstrated experiment uses a high-gain amplifier whose noise dominates the room-temperature environment, a passive detector of amplifier noise would actually outperform the quantum receiver in this exact setup; translating the quantum advantage into a practical radar therefore requires lower-gain, quantum-limited amplification or direct photon counting, a parameter regime the paper i","The methodological template of linear quadrature recording plus post-processed phase conjugation could be applied to other continuous-variable sensors, where the same covariance-based receiver could be tuned to different frequency bands without changing the digital post-processing."],"forward_implications":["A microwave quantum-illumination radar with a digital phase-conjugate receiver can outperform a symmetric classically correlated noise radar by up to 3 dB in raw SNR at the same power, bandwidth, and signal path.","If ideal idler photon-number detection is available, the same protocol would beat coherent heterodyne illumination by up to 4 dB and coherent homodyne illumination by about 1 dB at low photon numbers, where entanglement is present.","Because all receiver operations are done in post-processing, the scheme avoids the idler-storage time that limits analog photodetection receivers, so the usable radar range is not constrained by memory.","At less than one signal photon per mode, the protocol operates in a regime suitable for non-invasive scanning and short-range low-power radar, assuming the amplifier-noise challenge is handled."],"supporting_citations":[{"why":"supplies the Gaussian-state quantum illumination theory and the entanglement-based SNR benchmark the experiment seeks to realize","marker":"[12]"},{"why":"proposes microwave quantum illumination with a Josephson source and defines the target-detection protocol implemented here","marker":"[23]"},{"why":"provides the phase-conjugate receiver design and the SNR equations used for the digital receiver","marker":"[22]"},{"why":"previous microwave quantum-enhanced noise radar experiment used as the symmetric classical benchmark for comparison","marker":"[20]"},{"why":"provides the continuous-variable inseparability criterion used to verify the JPC output is entangled","marker":"[33]"},{"why":"the Josephson ring modulator device on which the Josephson parametric converter source is based","marker":"[24]"},{"why":"provides the linear-detector covariance reconstruction method used to turn digitized quadratures into mode statistics","marker":"[30]"}],"fun_headline_variants":["Microwave quantum illumination: 3 dB over classical noise radar","Quantum illumination at microwaves: digital receiver beats noise radar","Josephson converter enables microwave quantum illumination advantage","Quantum illumination radar: 3 dB gain with digital receiver","Microwave quantum illumination outperforms noise radar by 3 dB"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The headline quantum advantage over coherent-state homodyne detection rests on the assumption that subtracting the measured idler detector noise and vacuum contribution reproduces exactly what an ideal photon-number-resolving idler detector would see, even though no such detector was used.","fun_headline_variants_meta":{"raw":{"variants":["Microwave quantum illumination: 3 dB over classical noise radar","Quantum illumination at microwaves: digital receiver beats noise radar","Josephson converter enables microwave quantum illumination advantage","Quantum illumination radar: 3 dB gain with digital receiver","Microwave quantum illumination outperforms noise radar by 3 dB"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000719,"raw_usage":{"total_tokens":3223,"prompt_tokens":932,"completion_tokens":2291,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":548,"completion_tokens_details":{"reasoning_tokens":2211}},"tokens_in":548,"tokens_out":2291,"duration_ms":17128,"temperature":1.0,"reasoning_tokens":2211,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:25:51.995406+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Replace the calibrated idler post-processing with a real microwave photon-number-resolving detector at the JPC output and measure the SNR for the same target at $N_S<0.4$ photons per mode; if the realized advantage over coherent-state homodyne detection does not reach the predicted ~1 dB (or the heterodyne advantage does not reach ~4 dB), the simulated idler calibration is the point of failure.","supporting_citations":[{"cited_title":"Erkmen, Vittorio Giovannetti, Saikat Guha, Seth Lloyd, Lorenzo Maccone, Stefano Pirandola, Jeffrey H","cited_arxiv_id":null,"evidence_quote":"supplies the Gaussian-state quantum illumination theory and the entanglement-based SNR benchmark the experiment seeks to realize"},{"cited_title":"Shapiro, and Stefano Pirandola, Microwave quantum illumination","cited_arxiv_id":null,"evidence_quote":"proposes microwave quantum illumination with a Josephson source and defines the target-detection protocol implemented here"},{"cited_title":"Erkmen, Gaussian-state quantum- illumination receivers for target detection","cited_arxiv_id":null,"evidence_quote":"provides the phase-conjugate receiver design and the SNR equations used for the digital receiver"},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"previous microwave quantum-enhanced noise radar experiment used as the symmetric classical benchmark for comparison"},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"provides the continuous-variable inseparability criterion used to verify the JPC output is entangled"},{"cited_title":"Bergeal, R","cited_arxiv_id":null,"evidence_quote":"the Josephson ring modulator device on which the Josephson parametric converter source is based"},{"cited_title":"Eichler, D","cited_arxiv_id":null,"evidence_quote":"provides the linear-detector covariance reconstruction method used to turn digitized quadratures into mode statistics"}],"review_version":1}