{"id":"ffa65c2d-9c82-43da-8f2b-8e42456ea2d2","arxiv_id":"2411.15135","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Demonstrates continuous automatic polarization channel stabilization using heterodyne-detected dim wavelength-multiplexed references, maintaining entanglement fidelity 0.94 ± 0.03 over 30 hours on a metropolitan quantum network.","lead":"A team built an automatic polarization stabilizer that uses dim laser reference signals and fast heterodyne detection to continuously correct fiber polarization drift for quantum networking. It kept entanglement distribution stable with an average relative fidelity of 0.94 over more than 30 hours on a metropolitan fiber network.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 30-hr field result reports only fidelity relative to the first measured state, not absolute fidelity to a Bell state; the claim of 'high-fidelity entanglement distribution' is therefore not established.","rationale":"The reader's weakest assumption was polarization-mode dispersion: that the reference wavelengths and quantum wavelengths experience nearly the same polarization transformation. That is a legitimate concern, and the paper's own Figs. 9–10 show growing residual tilt with larger channel separation. However, the deployed field run used only a 2-channel separation over a few km, and the measured residual effect is small, so PMD is an empirical limitation rather than a direct threat to the central demonstration. The more load-bearing gap is the meaning of the headline fidelity number. The abstract claims 'continuous high-fidelity entanglement distribution' with 'average relative fidelity of 0.94±0.03'. Section 3.4 makes clear that this is the relative fidelity between the first estimated state and successive states, not fidelity to an ideal Bell state. A stable but noisy or misaligned channel could produce exactly this number while providing no usable entanglement. The low per-setting counts add statistical weight to the concern: without credible intervals, the ±0.03 is only the spread across tomographies, not the uncertainty of each fidelity estimate. This does not invalidate the paper's genuine contribution—continuous automatic polarization channel stabilization is demonstrated as a stability result—but it means the 'high-fidelity entanglement distribution' claim is not yet supported. The reader's verdict of CONDITIONAL remains appropriate; the condition should include reporting absolute fidelity or an entanglement witness for the 30-hr run. No change in verdict is needed, hence UNCHANGED.","tokens_in":24946,"tokens_out":9196,"duration_ms":89319,"concrete_test":"Ask the authors for the raw coincidence timestamp lists from the 30-hr run, or re-analyze the data if released. Reconstruct each of the 95 two-qubit states with the same Bayesian method, then compute (i) fidelity to the ideal |Φ+> state, (ii) fidelity to the best Bell state under local unitary rotation to remove any static polarization frame, and (iii) a CHSH-Bell or entanglement-witness statistic from the same coincidence data. If the median absolute fidelity to the best Bell state is below 0.9 or the witness is not violated across the run, the 'high-fidelity entanglement distribution' claim fails; if it is at or above 0.9, the concern is resolved and relative fidelity was a conservative proxy.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 3.4 reports only the 'relative generalized Ulhmann fidelity' between successive reconstructed states and between the first and subsequent states (Fig. 12). The headline number 0.94±0.03 is thus a measure of temporal repeatability, not of closeness to an entangled target. Two identical mixed states, even the maximally mixed state, have fidelity 1, so high relative fidelity is compatible with zero entanglement. The abstract's 'high-fidelity entanglement distribution' is unsupported unless an absolute fidelity to a Bell state, or an entanglement witness, is reported for the same run. The low statistics make this gap worse: the paper states an average of 310 coincidence counts per 36-setting tomography, roughly 8.6 counts per setting. Per-tomography statistical uncertainty is likely several percent, yet no error bars or credible intervals are shown for Fig. 12. The entanglement-assisted process tomography in Sec. 3.3 also uses relative process fidelity to the first Choi state, so it likewise establishes stability rather than absolute process quality. This is a correctness risk for the central headline claim, not a mere presentation issue.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a method for automatic polarization compensation (APC) for quantum networks, based on heterodyne detection of dim, wavelength-multiplexed reference signals and a multi-axis PID control algorithm with a slope-sign variation for extended range. The authors characterize the method with classical test-signal experiments over deployed metropolitan fiber, with entanglement-assisted process tomography, and with a 30-hour entanglement distribution run in which they report an average relative state fidelity of 0.94±0.03. The central claims are that this enables continuous, high-bandwidth, low-noise automatic polarization channel stabilization with no downtime and that the method is suitable for high-fidelity entanglement distribution.","tokens_in":25136,"tokens_out":3594,"duration_ms":35271,"significance":"If the claims are properly supported, the work is significant for deployed quantum networks: it offers a path to continuous polarization stabilization without time-multiplexing downtime, using a self-contained geometric control approach and a detailed experimental characterization over real in-ground fiber. The paper is strong in its treatment of design trade-offs, in the careful physical modeling of the control loop, and in providing extensive characterization of polarization drift in the field. The simulations and experiments are described in enough detail to be reproduced by a capable group, though no data or code are released. The main reservation is that the headline 'high-fidelity entanglement distribution' is supported only by relative fidelity metrics, which are stability measures, not measures of closeness to an entangled target; this must be corrected.","major_comments":[{"comment":"The abstract's claim of 'continuous high-fidelity entanglement distribution' is not established by the reported metric. The 0.94±0.03 value is the 'relative generalized Uhlmann fidelity' between the first estimated state and successive states (Fig. 12), not the fidelity to the intended Bell state. Two identical mixed states, including the maximally mixed state, have relative fidelity 1, so high relative fidelity is compatible with zero entanglement. The authors should either report absolute fidelity to the expected Bell state (or an entanglement witness) for the same 30-hour run, or rephrase the claim as stability of the reconstructed state over time.","section":"Abstract; Sec. 3.4, Fig. 12"},{"comment":"The statistical uncertainty of the plotted fidelities is not quantified. The text states an average of 310 coincidence counts per 36-setting tomography, i.e., about 8.6 counts per setting; the resulting per-tomography uncertainty is likely at the few-percent level, comparable to the observed drift in Fig. 12. Without error bars or credible intervals, it is difficult to separate genuine APC-induced drift from shot noise. The paper should provide confidence regions for the fidelities, for example using the Bayesian tomography methods already cited (Refs. [51,52]).","section":"Sec. 3.4, Fig. 12"},{"comment":"The claim of 'newly developed multi-axis (non-)linear control algorithm(s) for complete polarization channel stabilization with no downtime' conflates simulation with experiment. The slope-sign extended-range algorithm is demonstrated only in simulation (Figs. 3 and 4); the 30-hour experiment used the base algorithm without the slope sign, which has a π-range limitation and in fact relocked once (Sec. 3.4). The authors should clearly state that the no-downtime, complete-range property is a predicted capability supported by simulation, and should report the duration and effect of the observed relock on the entanglement distribution.","section":"Sec. 2.1; Sec. 3.4"},{"comment":"The validity of using reference wavelengths a few 100-GHz channels away from the quantum signal rests on the assumption that polarization-mode dispersion does not substantially decorrelate the two wavelengths. The tests in Sec. 3.2 and Fig. 10 are qualitative: statements such as 'slightly increased polarization noise', 'a slight tilt', and 'sufficiently insensitive' are not backed by numerical metrics. To support the PMD-insensitivity claim, the authors should provide quantitative residual-error measures (for example, RMS deviation from the setpoint or extinction-ratio stability over time) for the test signals at each wavelength separation, with uncertainties.","section":"Sec. 3.2; Sec. 4, PMD discussion"}],"minor_comments":[{"comment":"The phrase 'PMD has only a minor affect' should be 'minor effect' in both occurrences.","section":"Sec. 3.2; Sec. 4"},{"comment":"The name 'Ulhmann' is misspelled; the correct spelling is 'Uhlmann' (also in the text accompanying Fig. 12).","section":"Eq. (11); Sec. 3.4"},{"comment":"The data availability statement says the data are not publicly available; for a paper whose central claim is a 30-hour field demonstration, releasing at least the processed tomography data and analysis code would materially strengthen reproducibility and allow independent verification of the relative-fidelity statistics.","section":"Data Availability Statement"},{"comment":"The notation 'PUQWP1UHWP1UPCMUFSUCh' is difficult to parse; please clarify the order of the matrix products and define each symbol consistently, for example by stating that the transformations act in sequence from right to left.","section":"Eqs. (1)-(2)"}],"recommendation":"major_revision","confidential_remarks":"The paper fits the scope of the journal and the experimental work appears substantial. The main concern is the overstatement of 'high-fidelity entanglement distribution' based on relative fidelity; this is correctable by reanalysis or rewording. I see no evidence of problematic citation practices; the self-citations are to prior work by the same group that is directly relevant. The lack of public data is a limitation but not, by itself, a blocker if the claims are revised appropriately."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The core thing to know: this is a real engineering advance. The combination of dim WDM reference tones, heterodyne detection, and three-axis PID control for full channel stabilization is new in the cited literature, and it does address the up-time/noise/bandwidth trade-off better than time-multiplexed bright references or single-SOP stabilizers. The geometric control argument is coherent; the slope-sign extension is a sensible fix for the pi-range wrap; and the classical tracking data, including 1-Hz loop tracking of 10-Hz drift and the PMD sensitivity study across 2-10 channel spacings, are believable and useful.\n\nThe soft spots are real but mostly presentation-level. The stress-test note is right: the 30-hour headline number, 0.94±0.03, is a relative fidelity between tomographically estimated states, not an absolute fidelity to the intended Bell state. Two identical mixed states, even maximally mixed ones, have relative fidelity 1. The paper never reports an absolute fidelity, an entanglement witness, or concurrence for the 30-hour run. With only ~310 coincidences per 36-setting tomography, per-setting statistical error is several percent, and Fig. 12 has no error bars. So the abstract's \"high-fidelity entanglement distribution\" is not established; what is established is that the polarization channel stayed repeatable. That is still valuable, but the claim should be reworded.\n\nThe second soft spot: the \"complete stabilization\" algorithm with slope sign was only simulated. The 30-hour field run used the range-limited base PID and relocked once when a squeezer hit its range. So the demonstrated system had a downtime event, even if the method can in principle avoid it. The authors were candid about this in the text; the abstract is the problem.\n\nThird, no public data or code, and the classical tracking plots lack quantified error bars. That limits independent scrutiny. The no-APC comparison in the appendix uses a different 5-km spool, so it is not a controlled baseline. These are all fixable.\n\nI don't think the central method is flawed. The PMD assumption—reference wavelengths and quantum wavelength see nearly the same transformation—is empirically tested at a few km and several channel spacings, and the paper acknowledges the limit. The control derivation is self-contained, and no fitted value is folded into the reported fidelities.\n\nBottom line: worth a serious referee and likely worth publishing after the claims are aligned with the data. If I were handling it, I'd require an absolute fidelity or an explicit \"stability\" framing, error bars on the tomography, and ideally data/code release.","headline":"A genuinely useful APC architecture, honestly described in most places, but the abstract's 'high-fidelity entanglement distribution' overstates a stability demonstration.","tokens_in":25695,"tokens_out":2621,"would_cite":true,"duration_ms":26277,"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":"This paper claims that heterodyne detection of dim wavelength-multiplexed reference beams can stabilize the complete polarization channel of deployed fiber continuously, as demonstrated by 30 hours of metropolitan entanglement…","keywords":["automatic polarization compensation","heterodyne detection","polarization channel stabilization","entanglement distribution","metropolitan quantum network","polarization-mode dispersion","fiber squeezer","PID control"],"falsifier":"Take a deployed fiber longer than 10 km (or a spool under strong thermal gradients), place a test signal 10 or more 100-GHz channels away from the references, close the APC loop on the references, and monitor the stabilized test-signal Stokes drift or single-photon process fidelity. If the residual drift or fidelity loss exceeds the few-percent level observed here at 3–5 km, the PMD assumption behind the correction is falsified.","tokens_in":24722,"feed_emoji":"⚛️","tokens_out":10185,"duration_ms":87641,"temperature":0.7,"pith_summary":"Deployed optical fiber continuously scrambles the polarization of transmitted light, and quantum networks that encode information in polarization need to undo that scrambling in real time. This paper claims that a continuous, high-bandwidth, fully automatic polarization stabilizer can be built by transmitting two dim classical reference beams alongside the quantum signal, detecting them with heterodyne receivers, and driving three fiber squeezers with a three-loop control algorithm that stabilizes the complete polarization channel rather than a single polarization state. The authors demonstrate the method on a metropolitan fiber network, distributing polarization-entangled photons with an average relative fidelity of $0.94\\pm 0.03$ for more than 30 hours with no downtime and no measurable added noise from the reference beams. If the claim holds, it removes a central practical obstacle to operating polarization-encoded quantum links on real-world fiber infrastructure.","feed_headline":"Fiber polarization locked for 30 hours on city quantum network","feed_subtitle":"Dim reference beams, heterodyne receivers, and three PID loops hold entanglement fidelity at 0.94 with zero downtime.","key_machinery":"The central mechanism is a three-loop control system built on Poincaré-sphere geometry: a variable wave-plate rotates polarization along a great circle whose axis is the wave-plate's eigen-basis, so to control a wave-plate in basis 1 one measures a reference polarization in basis 3 along basis 2, with bases 1, 2, and 3 mutually unbiased. Two dim reference beams, one right-circular (R) and one vertical (V), are wavelength-multiplexed with the quantum signal; heterodyne detection of their projections onto the D, H, and R bases provides low-noise, high-bandwidth error signals even though the reference power is only about $-50$ dBm. Three independent PID loops actuate piezo fiber squeezers in the H/V, D/A, and R/L bases, and a finite-difference slope-sign correction to the error input lets the linear controller wrap around the Poincaré sphere, overcoming the $\\pi$ retardance range limit of a single loop.","core_discovery":"The central discovery is that heterodyne detection of dim, wavelength-multiplexed reference signals can support continuous, high-bandwidth automatic polarization channel stabilization with 100% uptime. Two reference signals—one right-circularly polarized, one vertically polarized and frequency-shifted by 200 MHz—travel through the fiber with the quantum signal. At the receiver they are split and projected onto three measurement bases (diagonal, horizontal, and circular) through waveplates and polarizers, and each projection is measured by a balanced heterodyne receiver. The filtered RF power of each projection becomes the error signal for an independent PID controller, and each controller drives a piezo fiber squeezer whose retardance axis is in a mutually unbiased basis (H/V, D/A, R/L). Because the three loops are nearly orthogonal at the operating point, together they lock the full polarization transformation of the channel for any input polarization, not just one state of polarization. A slope-sign modification to the error signal extends control over the full Poincaré sphere, avoiding the $\\pi$-range limit of ordinary linear control. The method is verified with classical test signals, entanglement-assisted process tomography over a 5-km fiber spool (average relative process fidelity $0.96\\pm 0.01$ successive, $0.94\\pm 0.03$ versus first), and a metropolitan entanglement distribution run with two independent APCs operating continuously for over 30 hours (average relative state fidelity $0.96\\pm 0.02$ successive, $0.94\\pm 0.03$ versus first).","pith_inferences":["The residual fidelity drift (0.94 versus first over 30 h) appears dominated by calibration drift of the passive measurement optics (3–5° waveplate shift per day) and by PMD, not by the control loop itself; automated recalibration or temperature stabilization of the receiver could push long-term fidelity higher. (My inference from the paper's logged calibration observations.)","The PMD sensitivity measured here (slight residual tilt at 5–10 channel separation over about 3 km) suggests that for longer fibers the two-reference scheme would need either closer channel spacing or a pair of references straddling the signal wavelength with averaged corrections—an extension the authors hypothesize but do not demonstrate.","Because the control algorithm only relies on mutual unbiasedness of three measurement bases, the same three-loop PID structure could be implemented with liquid-crystal or integrated-optic waveplates, trading speed and loss according to the actuator choice.","The demonstration of 100 Hz tracking under induced drift suggests the method's bandwidth headroom is roughly two orders of magnitude above the drift rates observed on in-ground metropolitan fiber, leaving margin for aerial or more exposed fiber."],"forward_implications":["Polarization-encoded quantum links over deployed fiber can run continuously at stable fidelity, with no time-multiplexing downtime and no reliance on application-specific feedback such as quantum bit error rates.","The dim (about $-50$ dBm) references add no measurable photon-count noise to single-photon quantum signals after a single 100-GHz DWDM filter, so the method is compatible with coexistence on standard telecom infrastructure.","Control bandwidth can be raised to track faster drifts: with a 1-kHz PID integration bandwidth the APC fully tracked induced 10-Hz polarization oscillations, and the same architecture can accept faster actuators.","The method generalizes beyond single-mode fiber to free-space links and to classical or continuous-variable channels, where the reference-power constraints are more relaxed."],"supporting_citations":[{"why":"Establishes heterodyne detection as a polarization-control sensing approach, the detection choice this paper builds on.","marker":"[11]"},{"why":"Show that multiple polarization references can stabilize the complete polarization channel, the goal this paper extends to continuous high-bandwidth operation.","marker":"[23,24]"},{"why":"Documents the π-range limitation of linear polarization control and the switching-retarder endless-control idea; the paper's slope-sign variation is designed to overcome exactly this limitation.","marker":"[6]"},{"why":"Quantifies Raman noise generated by a bright reference near a quantum signal, motivating the use of dim reference signals.","marker":"[25]"},{"why":"Establishes the sensitivity of heterodyne spectrometers that makes dim-reference detection feasible.","marker":"[35]"},{"why":"Provides the broadband C+L-band polarization-entangled photon source used in the entanglement demonstrations.","marker":"[41]"},{"why":"Supplies the Bayesian ancilla-assisted process tomography analysis used to measure relative process fidelity.","marker":"[47]"},{"why":"Provides Bayesian quantum state estimation used to compute relative state fidelities from tomography.","marker":"[51,52]"}],"fun_headline_variants":["Heterodyne APC locks fiber polarization for 30-hour city run","Continuous polarization control on quantum network, no downtime","Dim reference beams stabilize quantum link for 30 hours","City-scale quantum network keeps fidelity 0.94 with auto-APC"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire correction rests on the assumption that the fiber's polarization transformation at the reference wavelengths is effectively the same as at the quantum-signal wavelength; if polarization-mode dispersion separates the two, the loop stabilizes a channel other than the one the quantum signal experiences.","fun_headline_variants_meta":{"raw":{"variants":["Heterodyne APC locks fiber polarization for 30-hour city run","Continuous polarization control on quantum network, no downtime","Dim reference beams stabilize quantum link for 30 hours","City-scale quantum network keeps fidelity 0.94 with auto-APC"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.0006,"raw_usage":{"total_tokens":2850,"prompt_tokens":1040,"completion_tokens":1810,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":656,"completion_tokens_details":{"reasoning_tokens":1739}},"tokens_in":656,"tokens_out":1810,"duration_ms":14464,"temperature":1.0,"reasoning_tokens":1739,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T14:27:23.409856+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a deployed fiber longer than 10 km (or a spool under strong thermal gradients), place a test signal 10 or more 100-GHz channels away from the references, close the APC loop on the references, and monitor the stabilized test-signal Stokes drift or single-photon process fidelity. If the residual drift or fidelity loss exceeds the few-percent level observed here at 3–5 km, the PMD assumption behind the correction is falsified.","supporting_citations":[{"cited_title":"Optical polarisation control utilising an optical heterodyne detection scheme,","cited_arxiv_id":null,"evidence_quote":"Establishes heterodyne detection as a polarization-control sensing approach, the detection choice this paper builds on."},{"cited_title":"Polarization stabilization in optical communications systems,","cited_arxiv_id":null,"evidence_quote":"Documents the π-range limitation of linear polarization control and the switching-retarder endless-control idea; the paper's slope-sign variation is designed to overcome exactly this limitation."},{"cited_title":"Coexistent quantum channel characterization using spectrally resolved Bayesian quantum process tomography,","cited_arxiv_id":null,"evidence_quote":"Quantifies Raman noise generated by a bright reference near a quantum signal, motivating the use of dim reference signals."},{"cited_title":"Heterodyne spectrometer sensitivity limit for quantum networking,","cited_arxiv_id":null,"evidence_quote":"Establishes the sensitivity of heterodyne spectrometers that makes dim-reference detection feasible."},{"cited_title":"Broadband polarization-entangled source for c+l-band flex-grid quantum networks,","cited_arxiv_id":null,"evidence_quote":"Provides the broadband C+L-band polarization-entangled photon source used in the entanglement demonstrations."},{"cited_title":"Deployed quantum link characterization via Bayesian ancilla-assisted process tomography","cited_arxiv_id":"2410.00892","evidence_quote":"Supplies the Bayesian ancilla-assisted process tomography analysis used to measure relative process fidelity."}],"review_version":1}