{"id":"ce5cd214-7e19-480a-a25b-007ae430bdfb","arxiv_id":"2508.11234","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Ternary ADCs with joint-pilot-and-data estimation can mitigate channel estimation degradation in massive MIMO under mild conditions without extra pilots.","lead":"This paper proposes using ternary ADCs in massive MIMO systems to cut power for AIoT sensing, estimating channels with a joint-pilot-and-data scheme. It claims this reduces the accuracy loss from coarse quantization without added pilot overhead.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'mild conditions' for JPD gain are unspecified; they likely hinge on reliable data detection, which may not hold at low SNR typical of AIoT.","rationale":"The reader's verdict is UNVERDICTED because the abstract does not provide enough detail to verify the central claim. My stress-test focuses on the same gap: the unspecified 'mild conditions.' This concern is load-bearing because the validity of the JPD scheme depends on those conditions. Without them, the claim is unfalsifiable as stated. The proposed concrete test would settle the issue by checking whether the conditions hold in a realistic AIoT regime. Since even if the test passes, the abstract still requires full-text verification, the verdict remains UNVERDICTED until such evidence is provided. Thus no verdict change is recommended.","tokens_in":691,"tokens_out":2531,"duration_ms":31491,"concrete_test":"Locate the exact statement of the 'mild conditions' in the full text (likely a theorem, e.g., Theorem 1). Then run a Monte Carlo simulation for a low-power AIoT scenario: M=16 antennas, SNR=5 dB, QPSK, unknown noise variance, and a standardized channel model (e.g., 3GPP TR 38.901 UMi). Compare the MSE and SER of pilot-only estimation versus JPD with the proposed EM/VI-EM estimators. If JPD does not outperform pilot-only by at least 1 dB (or does not approach the unquantized bound within 3 dB), then the central claim fails in the target regime. If the full text's stated conditions are not met by this scenario, the paper must explicitly acknowledge that the benefits are conditional and not universal.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that the joint-pilot-and-data (JPD) scheme 'effectively mitigates performance degradation due to coarse quantization effects under mild conditions, without necessitating additional pilot overhead.' The abstract never states what these mild conditions are. The JPD approach refines the channel estimate using detected data symbols; for this to help, the initial pilot-only detection must be reliable enough that the data symbols carry information about the channel rather than error propagation. If the 'mild conditions' correspond to high SNR, large antenna arrays, or favorable channel structure, the claim would not apply to the low-power AIoT regime that motivates the paper. In particular, AIoT devices typically operate at low SNR (often 0–10 dB) and with limited antenna counts; under those conditions, unreliable data detection could degrade the JPD estimate relative to pilot-only. The paper also claims to handle unknown noise variance, but the abstract does not say how. Without access to the full text, the proof of the 'analytical findings' cannot be checked, and the precise domain of validity is unknown. This is not an internal inconsistency, but a missing qualification that is load-bearing: if the conditions exclude the intended deployment scenario, the headline contribution falls away.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes the use of ternary ADCs (T-ADCs) in massive MIMO systems for low-power AIoT sensing, and addresses channel estimation under coarse quantization. The channel is first estimated via a pilot-aided scheme and then refined with a joint-pilot-and-data (JPD) approach. The authors claim analytical findings that JPD mitigates quantization-induced channel estimation degradation under mild conditions without additional pilot overhead, and they propose modified EM and variational inference EM estimators for deterministic and random channels, respectively. Simulation results in terms of MSE and SER are reported, including a scenario with unknown noise variance. The abstract does not provide the analytical equations, the exact conditions for the claims, or details of the estimators.","tokens_in":982,"tokens_out":2764,"duration_ms":32990,"significance":"If the claims hold, the paper could make a useful contribution to low-power massive MIMO by showing that coarse ternary quantization does not necessarily require extra pilot overhead when data symbols are reused. The comparison with parallel one-bit ADCs and the development of EM/VI estimators are concrete elements that could be of interest to the AIoT and massive MIMO communities. However, the abstract alone does not provide enough technical content to verify the analytical findings, the domain of validity of the 'mild conditions,' or the behavior of the estimators when the noise variance is unknown. The contribution is potentially valuable, but its significance cannot be fully assessed from the submitted abstract.","major_comments":[{"comment":"The central claim that the JPD scheme 'effectively mitigates performance degradation ... under mild conditions' is load-bearing, yet the abstract never states what these mild conditions are. The authors should specify them precisely: minimum SNR, number of antennas, channel coherence structure, and the reliability of the initial data detection. If the conditions are equivalent to 'data detection is already reliable,' the improvement over pilot-only estimation may be limited to a regime outside the low-power AIoT setting. Please state the conditions and provide a concrete test, e.g., a simulation at low SNR or short coherence time where JPD underperforms pilot-only estimation, to delineate the domain of validity.","section":"Abstract, 'under mild conditions'"},{"comment":"The abstract claims a realistic scenario with unknown noise variance is considered, but it does not explain how the proposed EM/VI estimators handle this. Are the noise variance and channel jointly estimated? Is the problem identifiable with ternary observations? Does the 'mild conditions' require any prior knowledge about the noise distribution? Please specify the observation model and the estimation procedure for this case.","section":"Abstract, 'noise variance is unknown'"},{"comment":"The abstract reports 'analytical findings' but provides no equations, theorem statements, or explicit assumptions. Without these, the reader cannot assess whether the JPD gain is derived for the actual ternary ADC model or relies on simplifying assumptions (e.g., Gaussian quantization noise, known thresholds, high-SNR approximations). For a journal submission, the analytical results should be stated with their assumptions and a clear domain of validity.","section":"Abstract, 'Analytical findings'"}],"minor_comments":[{"comment":"The phrase 'hardware-ideal counterpart, the parallel one-bit ADCs (PO-ADCs)' is confusing. Why is a parallel one-bit ADC configuration 'hardware-ideal'? This should be clarified, possibly by explaining the relationship between two-threshold ternary ADCs and an array of one-bit ADCs.","section":"Abstract, 'hardware-ideal counterpart'"},{"comment":"The baseline for the claim 'without necessitating additional pilot overhead' should be explicit: compared to a pilot-only scheme using the same number of pilots, or compared to a scheme that adds pilots to achieve the same accuracy? Clarify the comparison.","section":"Abstract, 'without necessitating additional pilot overhead'"},{"comment":"The abstract refers to a 'two-threshold ADC system' and 'ternary ADCs,' but does not define how the thresholds are set. Are the thresholds fixed, optimized, or adapted to the channel/SNR? The choice of thresholds is a free parameter that could affect the 'mild conditions' and should be described.","section":"Abstract, 'two-threshold ADC'"}],"recommendation":"uncertain","confidential_remarks":"I am reviewing from the abstract only; the full text was not provided. My uncertainty reflects missing evidence, not a detected error. If the full text contains the missing conditions, the major comments are addressable. The main risk is that the 'mild conditions' may exclude the low-SNR, low-complexity AIoT regime the paper targets."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Abstract-only read, so take this as provisional. The paper looks like a credible incremental step in a mature field: it moves from one-bit to ternary ADCs for massive MIMO channel estimation and adds a joint-pilot-and-data (JPD) refinement stage that doesn't cost extra pilots. The proposed modified EM and variational inference EM estimators are the kind of thing you'd expect from someone who knows the area. If the analytical claims hold, the practical outcome is a small but real power saving for AIoT sensing, which is worth having.\n\nWhat the paper does well: it includes the parallel one-bit ADC as a hardware-ideal comparison, and it tackles unknown noise variance, which many papers conveniently assume away. Those choices make the simulations more honest.\n\nThe soft spot is the unspecified 'mild conditions' for the JPD gain. That matters because JPD refinement only helps if the initial pilot-based detection is already reliable enough that the detected data symbols carry channel information rather than error propagation. If those conditions secretly mean high SNR or large antenna arrays, then the low-power AIoT regime the paper aims at is not covered. The stress-test note is right to flag this; it's not an internal inconsistency, just a missing qualification that is load-bearing. The abstract also gives no equations and no prior-art comparison, so I can't judge novelty or rigor from this distance.\n\nMy honest verdict: it's a plausible paper that deserves a real referee, not a desk reject. The referee should force the authors to state the exact conditions and show the proof that JPD doesn't require extra pilots. If that holds up, it's a solid addition to the low-resolution ADC literature.\n\nFor my own work, I wouldn't cite it until I've seen the full text. But I'd bring it to a reading group that cares about the boundary conditions on data-aided channel estimation.","headline":"A plausible, incremental extension of one-bit ADC work to ternary ADCs with a joint-pilot-and-data refinement step; the unspecified 'mild conditions' are the main soft spot, but this deserves referee time rather than a desk reject.","tokens_in":1373,"tokens_out":1997,"would_cite":false,"duration_ms":22439,"reading_group":"maybe","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 a joint-pilot-and-data refinement recovers channel estimation accuracy lost to coarse ternary quantization in massive MIMO, without additional pilot overhead, making low-power AIoT sensing feasible.","keywords":["ternary ADC","massive MIMO","channel estimation","joint pilot and data","AIoT sensing","expectation maximization","variational inference","low-power receiver"],"falsifier":"Run the JPD estimator against a pilot-only estimator on measured or simulated massive MIMO channels with strong spatial correlation and low signal-to-noise ratio, with unknown noise variance, using ternary ADCs; if the JPD mean square error does not beat the pilot-only baseline at equal pilot overhead, or requires more pilots to match it, the claim is refuted.","tokens_in":635,"feed_emoji":"📡","tokens_out":2479,"duration_ms":29740,"temperature":0.7,"pith_summary":"The paper aims to show that massive MIMO systems for AIoT sensing can use cheap, low-power ternary ADCs instead of high-resolution converters without paying a large penalty in channel estimation accuracy. It proposes a two-stage approach: first estimate the channel from pilots, then refine that estimate using data symbols in a joint-pilot-and-data (JPD) scheme. The central claim is that JPD mitigates quantization-induced degradation under mild conditions while adding no pilot overhead. If true, this would allow energy-efficient massive MIMO receivers to still support reliable sensing, a practical step toward greener AIoT.","feed_headline":"Ternary ADCs recover MIMO channel accuracy with zero extra pilots","feed_subtitle":"A joint pilot-and-data refinement cuts quantization loss in low-power massive MIMO for AIoT sensing.","key_machinery":"The central object is the joint-pilot-and-data (JPD) refinement scheme applied to a two-threshold ternary ADC system: an initial pilot-based channel estimate is refined with data symbols to compensate for coarse quantization. The argument is carried by modified expectation-maximization and variational-inference EM estimators that handle deterministic and random channels respectively, with the unknown-noise-variance case treated as a realistic test.","core_discovery":"For a massive MIMO uplink with ternary ADCs (one-bit comparators plus a zero threshold), the authors claim that channel estimation degraded by coarse quantization can be substantially recovered by a joint-pilot-and-data scheme. The refinement uses modified expectation-maximization estimators for deterministic channels and variational-inference EM estimators for random channels, and the analysis explicitly includes the hardware-ideal parallel one-bit ADC counterpart as well as the realistic case where noise variance is unknown at the receiver. Under what they call mild conditions, JPD mitigates quantization loss without requiring additional pilot resources. The paper asserts that simulations","pith_inferences":["If JPD works under the stated mild conditions, the same refinement principle may extend to other coarse quantizers and to other receiver tasks such as activity detection or localization where pilot overhead is costly.","A practical next step would be to characterize the 'mild conditions' explicitly in terms of threshold spacing, noise level, and channel coherence so that deployment decisions can be made without relying on simulations.","The unknown-noise-variance analysis hints that JPD could be paired with online noise-variance estimation, removing a hidden calibration burden on low-power IoT nodes.","One testable extension is to compare JPD's recovery gain against the additional pilots that a pilot-only scheme would need to reach the same accuracy, quantifying the effective pilot savings in realistic conditions."],"forward_implications":["Ternary ADCs become a credible low-power front end for massive MIMO sensing in AIoT scenarios.","JPD improves channel estimation quality without spending additional pilot resources, preserving spectral efficiency.","The proposed EM and variational-EM estimators are shown to reduce mean square error and symbol error rate relative to pilot-only baselines.","Explicit treatment of unknown noise variance moves the scheme closer to realistic receiver operation.","The combined results indicate a feasible path to greener AIoT sensing with reduced ADC power consumption."],"supporting_citations":[],"fun_headline_variants":["Ternary ADCs plus joint data refine MIMO channels with no extra pilots","JPD scheme fixes coarse ADC loss in massive MIMO without extra pilots","Zero-pilot refinement recovers MIMO sensing accuracy with ternary ADCs","Joint pilot+data estimation trims quantization loss in ternary ADC MIMO","Ternary ADC MIMO: refined channel estimation without extra pilots"],"cache_read_input_tokens":2816,"weakest_assumption_plain":"The central premise is that the 'mild conditions' under which joint-pilot-and-data refinement offsets ternary-quantization loss actually hold for realistic AIoT massive MIMO channels; the abstract does not specify these conditions, leaving open that they might be narrow or defined by the simulation setup.","fun_headline_variants_meta":{"raw":{"variants":["Ternary ADCs plus joint data refine MIMO channels with no extra pilots","JPD scheme fixes coarse ADC loss in massive MIMO without extra pilots","Zero-pilot refinement recovers MIMO sensing accuracy with ternary ADCs","Joint pilot+data estimation trims quantization loss in ternary ADC MIMO","Ternary ADC MIMO: refined channel estimation without extra pilots"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000468,"raw_usage":{"total_tokens":2172,"prompt_tokens":748,"completion_tokens":1424,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":492,"completion_tokens_details":{"reasoning_tokens":1345}},"tokens_in":492,"tokens_out":1424,"duration_ms":11809,"temperature":1.0,"reasoning_tokens":1345,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T20:02:24.141796+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the JPD estimator against a pilot-only estimator on measured or simulated massive MIMO channels with strong spatial correlation and low signal-to-noise ratio, with unknown noise variance, using ternary ADCs; if the JPD mean square error does not beat the pilot-only baseline at equal pilot overhead, or requires more pilots to match it, the claim is refuted.","supporting_citations":[],"review_version":1}