{"id":"dea3be97-712a-4fe9-9f57-b66d3ce3f47b","arxiv_id":"2507.21546","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"First long-distance free-space CVQKD field test: positive asymptotic secret key rates over 7 km and 9.6 km atmospheric channels with all-day operation.","lead":"Researchers demonstrated continuous-variable quantum key distribution over 7 km inland and 9.6 km maritime free-space links, the longest outdoor distances reported for this technique. The result suggests satellite-to-ground continuous-variable QKD may be feasible in daylight without spectral filtering.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central claim rests on asymptotic key rates only; no finite-size or composable security analysis is provided, so positive Eq. (2) rates may not establish that secure keys were actually distributed over the 7-km and 9.6-km links.","rationale":"I read the paper as claiming a milestone demonstration: long-distance free-space CVQKD over 7-km inland and 9.6-km maritime channels with Gaussian-modulated coherent states, producing positive secret key rates under real atmospheric conditions. For that claim to hold, the computed rates must be legitimate lower bounds on achievable secure key rates for the finite experimental data sets. The paper uses Eq. (1), an asymptotic Devetak-Winter expression, and Eq. (2), a grouping average, and explicitly labels the treatment as asymptotic. The authors' own conclusion identifies finite-size effects as a significant challenge. This is not an internal inconsistency, but it is the most load-bearing weakness: without finite-size corrections, the reported rates cannot be certified as actual secure key rates, and the milestone claim is therefore conditionally supported rather than firmly established. The reader's weakest assumption concerned residual fading excess noise after 0.2-dB binning. I regard that as a valid but less decisive concern because the reported residual fading noise values are tiny compared with the reported excess noise, so the grouping procedure appears fit for its stated purpose. A finite-size reanalysis, however, could in principle change the sign of the net rate, which would directly invalidate the headline claim. I agree with the reader's conditional verdict and recommend no change to it; the paper should be accepted only with access to data or a finite-size analysis, or with the claim appropriately qualified as asymptotic.","tokens_in":8300,"tokens_out":6116,"duration_ms":88880,"concrete_test":"Request from the authors the per-group symbol counts or the raw acquisition durations for the Table I experiments, then recompute the lowest-SNR 9.6-km maritime group with a finite-size CVQKD bound (e.g., Leverrier et al., PRA 81, 062343 (2010), or a composable bound from Leverrier, PRL 114, 070501 (2015)) at 95% confidence, using the grouped-channel security framework of Ruppert et al. (NJP 21, 123036 (2019)) instead of the asymptotic Eq. (2). If the finite-size lower bound on R is non-positive for any headline group, then the claim that secure keys were distributed at 9.6 km is not supported by the presented evidence.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim—secure key distribution over 7-km and 9.6-km free-space links—is supported only by asymptotic secret-key rates computed from Eq. (1) and aggregated by Eq. (2). The Devetak-Winter bound in Eq. (1) assumes N→∞ and an effectively i.i.d. channel under collective attacks, while the field-test section applies it to finite grouped data blocks without finite-size corrections. The authors explicitly concede that \"transitioning from the asymptotic regime to the finite-length regime poses a significant challenge,\" and no composable parameter-estimation bounds, finite-size key-rate formulas, or actually distilled key lengths are reported. This is quantitatively serious: the optimal groups in Table I have channel-loss probabilities P as low as 0.05% and secure-key-generation proportions αG as low as 5.88%, so the number of raw symbols per group over a realistic stable atmospheric window is far below what the 2.5 MHz symbol rate alone suggests, especially at SNR values near 0.006 with FER up to 90%. The reader's binning concern is real but secondary: the measured residual fading noise in Fig. 2b is about 10^-5 SNU, roughly three orders of magnitude below the reported excess noise values of 0.0035–0.0487 SNU, so 0.2-dB grouping appears adequate for approximate stationarity. The load-bearing gap is the leap from positive asymptotic rate to the statement that secure quantum secret keys were generated; without finite-size analysis, the reported rates are an upper-bound estimate, not a demonstrated secure key rate.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports field tests of Gaussian-modulated coherent-state CVQKD over a 7-km inland atmospheric link and a 9.6-km maritime atmospheric link, using transmitted local oscillator, dynamic polarization and phase compensation, transmission-efficiency monitoring, and grouping of raw data into 0.2-dB channel-loss bins. The authors compute asymptotic secret key rates from the Devetak-Winter bound, report positive rates for selected 'optimal' groups in Table I, and claim the first long-distance free-space CVQKD demonstration beyond the previous 1.6-km record. The manuscript explicitly acknowledges that the analysis is asymptotic and that finite-size security remains future work.","tokens_in":8623,"tokens_out":4406,"duration_ms":59727,"significance":"If the claims are accepted with appropriate security qualifications, the work would be a notable experimental advance: it extends outdoor free-space CVQKD from 1.6 km to 7 km and 9.6 km, demonstrates operation in daylight without additional spectral filtering or wavelength conversion, and uses measured noise and loss parameters rather than fitted key rates. The core reported quantities, positive asymptotic secret key rates from Eq. (2), are supported by the standard Devetak-Winter formula applied to measured parameters. The main weakness is that the manuscript's language repeatedly asserts actual 'secure key distribution,' whereas the evidence establishes only positive asymptotic rates under an i.i.d./stationary-channel assumption, with no finite-size or composable security analysis, no error bars, and no raw data. The grouping approximation appears reasonable for the reported data because the measured residual fading noise in Fig. 2b is orders of magnitude below the reported excess noise, but the absence of finite-size corrections and full data reporting is load-bearing for the central claim.","major_comments":[{"comment":"The central claim that secure keys were distributed over 7 km and 9.6 km is supported only by asymptotic rates computed from Eq. (1) and aggregated by Eq. (2). Eq. (1) is the N→∞ Devetak-Winter bound, and Eq. (2) applies it to finite grouped data blocks without finite-size corrections or composable security bounds. This is quantitatively serious: Table I lists group probabilities P as low as 0.05% and secure-key-generation proportions αG as low as 5.88%, so the number of raw symbols in many groups is far below what the 2.5-MHz symbol rate alone would suggest, especially at SNRs near 0.006 with FER up to 90%. The authors themselves state that 'transitioning from the asymptotic regime to the finite-length regime poses a significant challenge,' yet the abstract and introduction assert that secure quantum secret keys were demonstrated. The manuscript must either provide a finite-size/composable key-rate analysis with explicit block lengths and parameter-estimation confidence intervals, or clearly restrict all claims of 'secure key distribution' to asymptotic secret-key-rate estimation.","section":"Field-test results and Eq. (1)-(2)"},{"comment":"Table I reports only the 'optimal group' for each experiment and gives no uncertainties on the dynamical parameters (L, ε, SNR, R) and no information about the other groups used in Eq. (2). The 'achieved average secret key rates' of 0.0560-2.3545 bps over 7 km and 0.0379-0.4768 bps over 9.6 km therefore cannot be independently reconstructed or checked. To support the experimental claims, the authors should provide the full distribution of group rates, the number of raw key symbols per group, a description of how the average over Eq. (2) is computed, and error bars or confidence intervals reflecting measurement uncertainty and statistical fluctuations.","section":"Table I"},{"comment":"The grouping procedure in Eq. (2) assumes that within each 0.2-dB transmission-efficiency bin the channel is sufficiently stationary for the standard GMCS CVQKD security proof to apply. The measured fading excess noise in Fig. 2b is about 10^-5 SNU, roughly three orders of magnitude below the excess-noise values of 0.0035-0.0487 SNU in Table I, so for the reported runs this assumption appears plausible. However, the manuscript provides no formal criterion for choosing ΔT = 0.2 dB, no validation that residual fading noise is negligible for every run, and no finite-size parameter-estimation bounds within each group. Since the key rate is sensitive to excess noise, a quantitative argument (or at least a sensitivity analysis) is needed to justify the grouping procedure as a general method rather than a case-specific heuristic.","section":"Scheme description, Eq. (2), and Fig. 2b"}],"minor_comments":[{"comment":"The abstract claims 'high-rate' and 'secure quantum secret keys,' but the reported average rates are below 2.4 bps and the security is asymptotic. The wording should be qualified to match the actual evidence, e.g., 'positive asymptotic secret-key-rate estimates' and 'high-rate relative to previous free-space CVQKD demonstrations.'","section":"Abstract and Conclusion"},{"comment":"The reference lists for [32] and [33] appear corrupted or duplicated, with repeated author names and incomplete titles. These should be corrected to the standard bibliographic entries.","section":"References [32]-[33]"},{"comment":"In the second paragraph of 'Scheme description,' 'to significantly relive the extra fading excess noise' should read 'relieve,' and later in 'Implementation setup' 'the polarization states of the received pules' should read 'pulses.'","section":"Scheme description"},{"comment":"The fading-noise values in Fig. 2b are plotted for different 0.2-dB channel-loss intervals but no error bars or sample counts are shown; adding these would help assess the statistical significance of the difference between the two links.","section":"Fig. 2b"},{"comment":"The statement that the achieved distance is 'well beyond the atmosphere's effective thickness' is overstated, since 9.6 km is comparable to the cited ~10-km effective thickness and 7 km is below it. 'Approaching or exceeding' would be more accurate.","section":"Conclusion and discussion"}],"recommendation":"major_revision","confidential_remarks":"The paper is a strong experimental result in terms of distance and daylight operation, but the gap between the asymptotic analysis and the claimed 'secure key distribution' is the central issue. I would encourage the editor to require either a finite-size analysis or a careful rewriting of all security claims as asymptotic estimates before publication. The grouping concern is secondary given the measured fading-noise level, but the lack of raw data and the selective reporting of only optimal groups are important for reproducibility."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things about this paper. First, if the data is real, it is a genuine experimental milestone: 7 km inland and 9.6 km maritime free-space links for CVQKD, beating the previous outdoor record by roughly an order of magnitude, with all-day operation and no extra spectral filtering. Second, the paper's central claim—that secure keys were distributed—is not actually supported by what is presented. The key rates are asymptotic Devetak-Winter bounds; there is no finite-size analysis, no composable security argument, and no distilled key length. The authors concede this in the conclusion, which is honest, but it means the title and abstract run ahead of the evidence.\n\nWhat is actually new: the engineering. The channel-fluctuation-independent polarization and phase control, the per-pulse transmittance monitoring using a 12-bit ADC, and the two-stage ATP system are credible technical contributions. The measured fading noise in Fig. 2b is about 10^-5 SNU, orders of magnitude below the reported excess noise, so the 0.2 dB grouping to suppress fading noise looks adequate. That part of the stress-test concern is secondary.\n\nWhere the soft spots are: the paper reports only the optimal group per experiment in Table I, gives no error bars, and ships no raw data. The achieved average rates (0.056 to 2.35 bps over 7 km, 0.038 to 0.48 bps over 9.6 km) are presented without the distribution across groups, so a reader cannot independently assess how representative those numbers are. Also, the reference list has apparent errors: several entries, e.g. refs [27] and [30] and misattributed authors, look garbled. That is sloppy and undercuts trust, but it is fixable.\n\nThe real load-bearing issue is the leap from asymptotic positive rates to the statement that secure quantum secret keys were generated. In an experimental paper claiming a distance record, a finite-size analysis is not optional if you want to use the word \"secure\" without qualification. The authors know this and say so, which helps. The paper would be better framed as a demonstration of asymptotic key-rate feasibility over record distances.\n\nWho is this for? People working on free-space QKD and satellite links. It deserves a serious referee—the engineering is substantial and the distance record is important. The referee should ask for either a finite-size treatment or a carefully revised abstract that says asymptotic rates. I would not desk-reject this, but it needs major revision before publication.\n\nMy recommendation: send to peer review, but flag the finite-size gap prominently in the review.","headline":"A real record-distance free-space CVQKD field test with clever engineering, but the title and abstract overclaim secure key distribution when only asymptotic key rates are shown.","tokens_in":9230,"tokens_out":1942,"would_cite":true,"duration_ms":26433,"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":"Long-distance free-space quantum key distribution now works over a 9.6-km daylight link.","keywords":["continuous-variable quantum key distribution","free-space quantum communication","Gaussian-modulated coherent states","atmospheric channel","daylight QKD","maritime channel","excess noise suppression","satellite quantum cryptography"],"falsifier":"Recompute the grouped secret key rates from the same field data with the bin width changed to 0.1 dB and to 0.4 dB. If the estimated rates change substantially, or if the fading excess noise measured inside a 0.2 dB bin is not negligible compared with the roughly $10^{-5}$ shot-noise-unit averages reported in Fig. 2(b), the stationarity assumption behind Eq. (2) would be falsified and the claimed rates would need revision.","tokens_in":8108,"feed_emoji":"🔑","tokens_out":8196,"duration_ms":83241,"temperature":0.7,"pith_summary":"This paper reports the first long-distance free-space quantum key distribution using continuous-variable coherent states, over a 7-km inland link and a 9.6-km maritime link, in daylight and at night. The authors claim positive asymptotic secret key rates under real atmospheric conditions, with the best individual subchannel rate exceeding 700 bps, without wavelength conversion or extra spectral filtering. Their central technical move is to make atmospheric fading noise negligible by grouping raw key data into narrow 0.2 dB transmission-loss bins and applying per-bin parameter estimation and rate calculation. If correct, this extends free-space CVQKD by roughly half an order of magnitude beyond the previous 1.6-km outdoor demonstration. The demonstrated distance also exceeds the atmosphere's effective thickness, so the paper proposes the same all-day capability as a route to daylight satellite-based quantum cryptography.","feed_headline":"Quantum keys travel 9.6 km through open air in daylight","feed_subtitle":"Continuous-variable QKD passes the atmosphere's effective thickness, opening a daylight path toward satellite links.","key_machinery":"The load-bearing mechanism is a chain of channel-fluctuation-independent controls followed by grouped parameter estimation. Alice and Bob monitor each pulse's transmission efficiency in real time by sampling the 5% local-oscillator branch at 1 GS/s, then sort the sifted key data into 0.2 dB-wide loss intervals, the probability distribution of transmission efficiency (PDTE) grouping. Within a group the channel is treated as stationary, so the standard Gaussian-modulated coherent-state (GMCS) security proof applies and the total secret key rate is the probability-weighted sum of per-group rates. Two supporting controls carry the system: a polarization feedback that uses the ratio of leaked signal and local-oscillator light, making it independent of loss fluctuations, and two adjacent QPSK pilot pulses that calibrate the phase of each quantum signal. A coarse-fine acquisition, tracking, and pointing (ATP) stage keeps the link inside the regime where these corrections operate.","core_discovery":"The system prepares 1550-nm Gaussian-modulated coherent states, transmits them with a polarization-division and time-division multiplexed local oscillator, and detects one quadrature per pulse with homodyne detection. After each pulse's transmission efficiency is monitored in real time, the sifted raw keys are sorted into 0.2 dB loss groups; the secret key rate is then the probability-weighted sum of per-group rates from the reverse-reconciliation rate bound. The reported average secret key rates range from 0.0560 to 2.3545 bps over the 7-km inland link and from 0.0379 to 0.4768 bps over the 9.6-km maritime link, with the best subchannel rate exceeding 700 bps. The paper states this is the first long-distance free-space CVQKD demonstration and that the 9.6-km maritime distance exceeds the atmosphere's effective thickness, making daylight satellite-based quantum cryptography a proposed next application. A stated limitation is that the rates are asymptotic, not finite-size secure.","pith_inferences":["A direct test of the 0.2 dB binning assumption would be to recompute rates at 0.1 dB and 0.4 dB bin widths on the same field data; a rate that barely moves would support stationarity, while a strong dependence would mean residual fading noise is still present.","The maritime link's night-time difficulty points to fog and humidity, not the quantum protocol, as the practical availability limit; link-availability modeling would be a natural extension.","A satellite version of this architecture would face platform vibration, Doppler shifts, and faster pointing dynamics that a ground-to-ground field test cannot fully exercise, so the ATP design would need separate validation."],"forward_implications":["The previous outdoor distance ceiling of about 1.6 km is broken by roughly half an order of magnitude in a single field demonstration.","Daylight operation without wavelength conversion or extra spectral filtering removes a major practical obstacle to round-the-clock free-space quantum key distribution.","Because the system uses telecom-band coherent optical components, the same terminal could in principle interconnect with existing ground fiber quantum networks.","All reported rates are asymptotic; converting the demonstration into finite-size secure keys will require more accumulated data and an optimized grouping interval, which the authors identify as the next step."],"supporting_citations":[{"why":"The Gaussian-modulated coherent-state protocol with homodyne detection that the system implements.","marker":"[2]"},{"why":"Identifies the fading excess noise from atmospheric transmission-efficiency fluctuation, the central obstacle the field test suppresses.","marker":"[18]"},{"why":"Gives the grouping of raw key data by transmission-efficiency intervals that underlies Eq. (2).","marker":"[24]"},{"why":"The previous outdoor free-space CVQKD demonstration at up to 1.6 km, the record this experiment extends.","marker":"[30]"},{"why":"Supplies the reverse-reconciliation secret-key-rate bound used in Eq. (1).","marker":"[31]"},{"why":"Provides the multidimensional reconciliation used to distill keys at low signal-to-noise ratios.","marker":"[34]"}],"fun_headline_variants":["Quantum keys cross 9.6 km in daylight for first time","Daylight QKD sets 9.6-km maritime record","Free-space quantum keys pass atmosphere's effective thickness","CVQKD spans 9.6 km of open air in daylight","First daylight CVQKD link exceeds 9.6 km over sea"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The security analysis assumes that sorting received key data into 0.2 dB-wide loss groups makes the leftover noise from atmospheric fading negligible, so each group can be treated as a stationary channel for rate calculation; if that residual fading noise is not negligible, the reported secret key rates could be overestimated.","fun_headline_variants_meta":{"raw":{"variants":["Quantum keys cross 9.6 km in daylight for first time","Daylight QKD sets 9.6-km maritime record","Free-space quantum keys pass atmosphere's effective thickness","CVQKD spans 9.6 km of open air in daylight","First daylight CVQKD link exceeds 9.6 km over sea"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000784,"raw_usage":{"total_tokens":3464,"prompt_tokens":951,"completion_tokens":2513,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":567,"completion_tokens_details":{"reasoning_tokens":2424}},"tokens_in":567,"tokens_out":2513,"duration_ms":22475,"temperature":1.0,"reasoning_tokens":2424,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T12:38:06.463877+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the grouped secret key rates from the same field data with the bin width changed to 0.1 dB and to 0.4 dB. If the estimated rates change substantially, or if the fading excess noise measured inside a 0.2 dB bin is not negligible compared with the roughly $10^{-5}$ shot-noise-unit averages reported in Fig. 2(b), the stationarity assumption behind Eq. (2) would be falsified and the claimed rates would need revision.","supporting_citations":[{"cited_title":"Grosshans and P","cited_arxiv_id":null,"evidence_quote":"The Gaussian-modulated coherent-state protocol with homodyne detection that the system implements."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Identifies the fading excess noise from atmospheric transmission-efficiency fluctuation, the central obstacle the field test suppresses."},{"cited_title":"Ruppert, P","cited_arxiv_id":null,"evidence_quote":"Gives the grouping of raw key data by transmission-efficiency intervals that underlies Eq. (2)."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The previous outdoor free-space CVQKD demonstration at up to 1.6 km, the record this experiment extends."},{"cited_title":"Devetak and A","cited_arxiv_id":null,"evidence_quote":"Supplies the reverse-reconciliation secret-key-rate bound used in Eq. (1)."},{"cited_title":"Laudenbach, P","cited_arxiv_id":null,"evidence_quote":"Provides the multidimensional reconciliation used to distill keys at low signal-to-noise ratios."}],"review_version":1}