{"id":"e3d86bac-194c-42db-bcfb-252944240031","arxiv_id":"2501.08891","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Time-bin encoded quantum key distribution in the C-band is demonstrated over 50 m and 500 m horizontal free-space links with average secure key rates of 793 kbps and 40 kbps.","lead":"This experiment shows that quantum key distribution with time-bin encoding at telecom wavelengths works over short open-air links in a city, both by day and by night. The rates, 793 kbps at 50 meters and 40 kbps at 500 meters, suggest that fiber-style QKD equipment could be reused on free-space paths, easing connections between ground fiber networks and future satellite links.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'secure key rate' claim is not justified: the CW-laser source is not described as phase-randomized, so the decoy-state finite-key formula (Eq. 3) does not apply to the reported SKR values.","rationale":"I read the paper in good faith as a field demonstration whose central claim is practical viability, not a new security proof. The measured losses, visibilities, and turbulence parameters plausibly show that time-bin encoding in the C-band can survive 50 m and 500 m horizontal free-space links. However, the advertised quantitative outcome is a 'secure key rate', and that number is obtained from a finite-key decoy-state formula whose phase-randomization precondition is neither stated nor implemented in the described source. This is the most load-bearing point because the central claim would be materially weakened if the numbers are only raw rates. The concern is not that the experiment is fake; it is that the security conclusion is unsupported unless either the authors can point to a phase-randomization element omitted from the text, or the rates are recharacterized. The reader's weakest_assumption identifies exactly this issue; my read strengthens it by noting that a narrow-linewidth CW laser carved by an intensity modulator is not merely silent on phase randomization but, as described, actively maintains phase coherence between pulses. I therefore agree with the reader's conditional assessment and recommend no change to the verdict: the paper should be accepted only if the phase-randomization condition is demonstrated or the security claim is downgraded. The proposed interference test would settle the factual question quickly.","tokens_in":11600,"tokens_out":11136,"duration_ms":131183,"concrete_test":"Perform a first-order interference measurement between successive emitted pulses, e.g., send the attenuated source output into an unbalanced interferometer with a delay equal to the 1.68 ns pulse period, and record the fringe visibility over many pulse periods. If the visibility is significantly above zero, the source is phase-coherent across pulses and the phase-randomization assumption required by Eq. (3) is falsified; if the visibility averages to zero, the assumption is supported. An alternative decisive check is to re-derive the finite-key bound with the actual phase-coherent source state and verify whether it reduces to the decoy-state expression used in the paper.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract's central numbers are 'average secure key rate' of 793 kbps and 40 kbps. Those values are computed with Eq. (3), the finite-key bound for the three-state one-decoy BB84 protocol of refs. [20,29,34]. That security proof requires the source to emit phase-randomized weak coherent pulses, i.e., each pulse must be described as a Poisson mixture of Fock states. The source described in Sec. 2 and Suppl. 7.3 is a <100 kHz-linewidth CW laser carved by an intensity modulator into 595 MHz pulses; no phase modulator, gain-switched laser, or equivalent randomization element appears in the setup. With a <100 kHz linewidth, the coherence time is on the order of microseconds, far longer than the 1.68 ns pulse spacing, so successive pulses are phase-coherent or slowly drifting rather than uniformly random in phase. The 0/pi phases used for the X-basis are encoding values, not phase randomization. Without phase randomization, the decoy-state yield and phase-error bounds entering Eq. (3) are not valid; the quoted SKRs are at best raw key rates or unproven upper bounds. The physical demonstration of time-bin C-band propagation may still be sound, but the headline security claim does not follow as written.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports an experimental demonstration of time-bin encoding quantum key distribution in the C-band over 50 m and 500 m horizontal free-space links. The transmitter uses a continuous-wave laser carved by an intensity modulator at 595 MHz together with a free-space receiver that applies tip-tilt correction and a photonic integrated imbalanced Mach-Zehnder interferometer for the X basis. The authors report average visibilities of 94% and 85% and average secure key rates of 793 kbps and 40 kbps, computed with a finite-key formula for a three-state one-decoy BB84 protocol.","tokens_in":11811,"tokens_out":10034,"duration_ms":102586,"significance":"If the security claims were fully supported, the experiment would be a valuable step toward interoperable fiber/free-space QKD in the C-band, with the advantages of reduced solar background and turbulence compared with shorter wavelengths. The paper also provides useful engineering data: active beam compensation, measured scintillation indices, and channel losses for the two links. However, the headline secure-key-rate numbers depend on source and decoy-state assumptions that are not documented or implemented, and the manuscript does not report the QBER values or error bars needed to verify the finite-key calculation. The physical demonstration of time-bin interference over turbulent horizontal links is interesting, but the central claim of 'secure key rate' is not presently justified.","major_comments":[{"comment":"The finite-key formula in Eq. (3) is a decoy-state bound that assumes the source emits phase-randomized weak coherent pulses, i.e., a Poisson mixture of Fock states. The described source, a <100 kHz linewidth CW laser carved by an intensity modulator, emits phase-coherent pulses when the coherence time is far longer than the 1.68 ns pulse spacing. No phase modulator, gain-switched laser, or equivalent randomization element appears in Figure 1 or Supplementary 7.3. Without phase randomization, the decoy-state estimates for s_l_Z,0, s_l_Z,1, and the phase error rate entering Eq. (3) are not valid, and the values labeled 'secure key rate' are not established as secure. The authors must either implement and characterize phase randomization or withdraw the security interpretation of the reported rates.","section":"Section 2 and Supplementary 7.3, Eq. (3)"},{"comment":"The protocol is described as 'one-decoy' but the manuscript never explains how decoy states are generated or measured. The transmitter includes one intensity modulator, a VOA, and fixed attenuators; no random intensity modulation for signal and decoy states is mentioned. The secure key rate is computed solely from QBER_Z and QBER_X, but the bounds s_l_Z,0 and s_l_Z,1 in Eq. (3) require observed counts from at least two intensities. Without these decoy-state statistics, the SKR calculation is not reproducible and the use of the decoy-state formula is unsupported. Please provide the decoy implementation details and the measured signal/decoy yields, or remove the decoy-state-based finite-key formula.","section":"Section 2 and Supplementary 7.1, Eq. (3)"},{"comment":"The paper reports average SKR values but no measured QBER_Z, QBER_X, or error bars for any trial; visibility values also lack standard deviations or confidence intervals. Since the SKR is a nonlinear function of QBER_X and the visibility fluctuates over time (Fig. 2c), the reader cannot assess the statistical reliability of the quoted numbers or reproduce the finite-key calculation. A table listing per-trial parameters (nZ, QBER_Z, QBER_X, visibility, channel loss, integration time) is needed. In addition, Eq. (10) derives QBER_X directly from the raw interference visibility; for a multi-photon source the single-photon phase error rate used in Eq. (3) is not automatically equal to (1-V)/2. The manuscript should justify this substitution or use a proper phase-error estimation method.","section":"Section 3, Figure 2 and Supplementary 7.3"}],"minor_comments":[{"comment":"The open-loop versus closed-loop position errors (92 um vs 65 um) are cited as an improvement, but the standard deviations (53 um and 36 um) are large; a statistical confidence interval or test would strengthen the claim.","section":"Section 3"},{"comment":"The title mentions 'nighttime and daytime,' but the results do not explicitly separate day and night trials nor analyze the effect of ambient light on the measured rates; please clarify what the daytime and nighttime measurements show.","section":"Title and Section 3"},{"comment":"In Eq. (5), the symbol sigma is used both for the standard deviation and in expressions for the log-intensity variance, and the notation 'sigma^2/mu^2' is inconsistent with the definition of the scintillation index in Eq. (4); a notation cleanup would avoid ambiguity.","section":"Supplementary 7.2, Eq. (5)"},{"comment":"The statement that solar spectral irradiance at 1550 nm is about five times lower than at 800 nm lacks a specific citation or calculation; please add a reference or a brief derivation.","section":"Introduction"},{"comment":"The extrapolation 'up to 400 bps' for a 38 dB loss channel and '4 kbps in a 25 dB channel' is not backed by any model; please describe the assumptions used for these estimates.","section":"Section 4"},{"comment":"The SKR bar chart shows no error bars, yet the text quotes average SKRs with a range; adding error bars or a table of individual runs would make the variation explicit.","section":"Figure 2d"}],"recommendation":"reject","confidential_remarks":"The experimental effort, including the active-tip-tilt correction, the turbulence characterization, and the demonstration of stable time-bin interference over horizontal free-space links, is potentially valuable for the quantum communication community. However, the central 'secure key rate' claims are built on decoy-state and phase-randomization assumptions that are neither implemented nor described in the manuscript. The absence of a phase randomizing element and of any random intensity modulation for decoy states means that Eq. (3) cannot be applied as written. This is not a presentation issue; it affects the main quantitative results. Unless the authors can demonstrate (or add) the missing hardware and provide full decoy-state statistics, the paper would need to be reframed as a free-space channel characterization rather than a QKD field trial. Given the scope of the required change, I recommend rejection at this stage, though a substantially revised manuscript might be reconsidered."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, what you should know: the headline numbers (793 kbps and 40 kbps) are called average secure key rates, but the security analysis has a load-bearing gap. The source is a <100 kHz CW laser carved by an intensity modulator, and the paper never says the emitted pulses are phase-randomized. The finite-key formula in Eq. (3) comes from a decoy-state proof that assumes phase-randomized weak coherent pulses. A CW laser carved into 1.68 ns pulses has a coherence time of microseconds, so successive pulses are phase-coherent, not independently randomized. That means the decoy-state bounds on yields and phase errors don't apply, and the quoted SKRs are not proven secure key rates as written. The stress-test note has this right.\n\nThe experimental demonstration itself is plausible and worth credit. The combination of C-band time-bin encoding, a 595 MHz clock, a PIC-based imbalanced Mach-Zehnder decoder, and day/night operation over 50 m and 500 m horizontal free-space links is new. Measured visibilities (94% and 85%) and channel losses (7 dB and 16–17 dB) are consistent with a working system. The turbulence characterization is a nice addition, and the authors are honest about the short ranges and the role of undersized optics.\n\nThe other soft spots are minor in comparison: no QBER values, no error bars on SKR or visibility, and X-basis QBER is derived from visibility using a formula that assumes a symmetric interferometer. Those are easily fixed by reporting raw detection statistics and the security parameters used in Eq. (3) (epsilon_sec, epsilon_corr, feff). The phase randomization issue is more serious; it needs either a clear statement that the source is actively randomized (with a phase modulator or gain-switched design) or a rephrasing of the security claim as a transmission demonstration without the decoy-state security guarantee.\n\nWho benefits: readers interested in fiber-to-air interoperability will get a useful feasibility reference, but anyone citing the secure key rate should wait for a revised version. I'd send this to peer review because the core transmission result is credible and the security gap is fixable in revision, but I'd ask for the source characterization, raw data, and proper error analysis before acceptance. For now, I would not cite the SKR numbers in my own work.","headline":"The time-bin C-band free-space demonstration looks genuine, but the headline secure-key-rate numbers rest on a phase-randomization assumption the paper never states.","tokens_in":670,"tokens_out":673,"would_cite":false,"duration_ms":37317,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P94"],"pacs":["03.67.Dd","42.79.Sz"],"model":"deepseek-v4-flash","headline":"Time-bin encoded QKD in the C-band works over turbulent horizontal free-space links of 50 m and 500 m.","keywords":["quantum key distribution","time-bin encoding","free-space optics","C-band","turbulent atmosphere","finite key rates","BB84","photonic integrated circuit"],"falsifier":"Measure the relative phase of successive carved pulses at Alice's output with a fast unbalanced interferometer or heterodyne detection; if the phase differences are not uniformly random over $0$ to $2\\pi$, the source is not phase-randomized and the finite-key rates from Eq. (3) cannot be claimed as proven secure rates for this transmitter.","tokens_in":11362,"feed_emoji":"🔐","tokens_out":5213,"duration_ms":55476,"temperature":0.7,"pith_summary":"The paper reports field demonstrations of quantum key distribution that use time-bin-encoded qubits in the C-band (1558.98 nm) across horizontal free-space links of 50 m and 500 m in Florence. Average secure key rates reached 793 kbps and 40 kbps, with interferometric visibility of 94% and 85%, over several hours and in both daytime and nighttime conditions. The central claim is that this combination, time-bin encoding plus telecom C-band, is viable in turbulent air, making fiber-based and free-space QKD interoperable. If true, hybrid networks connecting fiber users, ground stations, and satellites could run on a single protocol and wavelength band instead of requiring conversion devices.","feed_headline":"Time-bin quantum keys cross 500 m of air at 40 kbps","feed_subtitle":"A fiber-compatible C-band setup hits 793 kbps at 50 m with 94% visibility, bridging fiber and free-space QKD.","key_machinery":"The central object is the time-bin qubit pair, with early and late pulses separated by 800 ps, generated at a 595 MHz rate by an intensity modulator carving a continuous-wave C-band laser. At the receiver, a 50:50 beam splitter implements Bob's random basis choice: one output goes directly to a superconducting nanowire single-photon detector for the Z basis, and the other goes into a photonic integrated circuit imbalanced Mach-Zehnder interferometer with an 800 ps delay for the X basis, from whose interference visibility the X-basis error rate is derived. The system also uses a 1310 nm beacon laser, a four-quadrant detector, and a PID-controlled deformable mirror to cancel beam wandering, plus a clock-over-air channel for time synchronization. Secure key lengths are extracted with the finite-key decoy-state formula for a three-state efficient BB84 protocol.","core_discovery":"The paper demonstrates that a time-bin encoded QKD protocol operating in the C-band can survive turbulent horizontal free-space channels. On a 50 m link with roughly 7 dB loss and a 500 m link with 16–17 dB loss, both in the weak-turbulence regime, the setup sustained average secure key rates of 793 kbps and 40 kbps, respectively. The authors attribute the improved performance to a 595 MHz qubit generation rate, coupling into single-mode fiber to reject background light, and a photonic integrated imbalanced Mach-Zehnder interferometer that keeps visibility stable at 94% and 85% over two-hour stretches. The paper frames this as evidence that time-bin encoding, normally considered fragile in free space because of phase instabilities, can be made practical with active tip-tilt beam correction and a telecom-wavelength link design.","pith_inferences":["If phase randomization of the emitted pulses were explicitly verified, this trial would be the first decoy-state-secure demonstration of time-bin C-band free-space QKD; the paper does not claim such verification.","The 500 m link's loss is dominated by undersized receiving optics rather than atmospheric attenuation, so scaling to kilometer-scale or slant-path links appears plausible with larger apertures, though the paper only proposes such upgrades.","The clock-over-air scheme with a reduced 145 kHz synchronization frequency for the longer link suggests a practical path to free-running alignment, at the cost of lower event-rate capacity that a full system design would need to budget."],"forward_implications":["A single transmitter and receiver design can serve both fiber and free-space legs of a hybrid quantum network without wavelength or encoding conversion.","At the measured 16 dB loss the 500 m link delivered 40 kbps, and the paper estimates roughly 400 bps at 38 dB loss and 4 kbps at 25 dB loss with optimized optics, indicating where the approach can operate as losses grow.","Active tip-tilt correction lowers the mean position error of the received beam, which stabilizes fiber coupling and interferometric visibility over multi-hour sessions.","Closing windows to suppress convective turbulence added about 3 dB of loss, so the reported rates include an extra attenuation that would not be present in a purpose-built outdoor terminal."],"supporting_citations":[{"why":"Supplies the security proof for the simplified three-state BB84 protocol used in the experiment.","marker":"[20]"},{"why":"Provides the earlier high-rate time-bin QKD implementation whose protocol structure this work adapts.","marker":"[29]"},{"why":"Introduces the decoy-state method that underpins the secure key length calculation.","marker":"[30]"},{"why":"Gives the finite-key secure key length formula, Eq. (3), used to convert measured QBERs into the reported key rates.","marker":"[34]"},{"why":"Is the earlier genuine time-bin free-space QKD demonstration over a turbulent channel that this work compares against and improves upon.","marker":"[26]"},{"why":"Reports the borosilicate-glass integrated photonic receiver, the imbalanced Mach-Zehnder interferometer chip used for X-basis detection.","marker":"[28]"},{"why":"Motivates the C-band choice by showing the wavelength dependence of turbulence-induced optical blurring.","marker":"[21]"},{"why":"Justifies telecom wavelengths for daytime operation via lower solar irradiance and reduced Rayleigh scattering compared to shorter wavelengths.","marker":"[23]"}],"fun_headline_variants":["Time-bin QKD in C-band: 793 kbps at 50 m, 40 kbps at 500 m","C-band time-bin QKD bridges fiber and free space","Time-bin encoding survives turbulent free-space links at 793 kbps","40 kbps secure key over 500 m of air with time-bin QKD"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The secure-key-rate calculation assumes the laser source emits pulses with random phases, as the decoy-state security proof requires, but the paper does not report checking this phase-randomization condition.","fun_headline_variants_meta":{"raw":{"variants":["Time-bin QKD in C-band: 793 kbps at 50 m, 40 kbps at 500 m","C-band time-bin QKD bridges fiber and free space","Time-bin encoding survives turbulent free-space links at 793 kbps","40 kbps secure key over 500 m of air with time-bin QKD"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000595,"raw_usage":{"total_tokens":2786,"prompt_tokens":944,"completion_tokens":1842,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":560,"completion_tokens_details":{"reasoning_tokens":1766}},"tokens_in":560,"tokens_out":1842,"duration_ms":12292,"temperature":1.0,"reasoning_tokens":1766,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T20:14:41.319530+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the relative phase of successive carved pulses at Alice's output with a fast unbalanced interferometer or heterodyne detection; if the phase differences are not uniformly random over $0$ to $2\\pi$, the source is not phase-randomized and the finite-key rates from Eq. (3) cannot be claimed as proven secure rates for this transmitter.","supporting_citations":[{"cited_title":"Security proof for a simplified Bennett-Brassard 1984 quantum-key-distribution protocol","cited_arxiv_id":null,"evidence_quote":"Supplies the security proof for the simplified three-state BB84 protocol used in the experiment."},{"cited_title":"Time-bin Encoded Pho- tons over Multi-mode Channels: Advances on Quantum Communications and Sensing Over Free- space Channels","cited_arxiv_id":null,"evidence_quote":"Provides the earlier high-rate time-bin QKD implementation whose protocol structure this work adapts."},{"cited_title":"De- coy state quantum key distribution","cited_arxiv_id":null,"evidence_quote":"Introduces the decoy-state method that underpins the secure key length calculation."},{"cited_title":"Secure Quantum Key Dis- tribution over 421 km of Optical Fiber","cited_arxiv_id":null,"evidence_quote":"Gives the finite-key secure key length formula, Eq. (3), used to convert measured QBERs into the reported key rates."},{"cited_title":"Quantum key distribution with a borosilicate-glass-matrix integrated photonic receiver","cited_arxiv_id":null,"evidence_quote":"Reports the borosilicate-glass integrated photonic receiver, the imbalanced Mach-Zehnder interferometer chip used for X-basis detection."},{"cited_title":"Long-distance free-space quantum key distribution in daylight towards inter- satellite communication","cited_arxiv_id":null,"evidence_quote":"Justifies telecom wavelengths for daytime operation via lower solar irradiance and reduced Rayleigh scattering compared to shorter wavelengths."}],"review_version":1}