{"id":"4d338367-3b02-4624-aa36-ca3a048f72d6","arxiv_id":"2606.28454","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":7.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"Con-focusing achieves the rank-one eigenbound in leading constant for LoS near-field MIMO power transfer by directing both apertures at the point from which they subtend equal angles, degenerating to steering for equal apertures.","lead":"This paper shows that in line-of-sight near-field MIMO with large apertures at both ends, traditional focusing is outperformed by far-field steering beyond a Fresnel number of 1.947 for equal apertures. It introduces con-focusing, where both apertures target a common point subtending equal angles, as the order-optimal strategy.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.3","headline":"Fresnel approximation error may invalidate leading-constant optimality claims precisely where Fresnel number is large","rationale":"The reader's weakest_assumption correctly identifies the Fresnel-regime assumption as central, but the load-bearing risk is not merely the assumption's existence but its accuracy inside the claimed operating regime. The proposed test directly quantifies whether the approximation error affects the headline scaling and optimality statements.","tokens_in":1764,"tokens_out":311,"duration_ms":46811,"concrete_test":"For equal apertures at Fresnel number F=10, evaluate the exact (non-Fresnel) power transfer integral for both focusing and con-focusing using the true Euclidean distance in the phase factor; if the ratio of the two gains deviates from the closed-form prediction by more than 1.5 dB, the leading-constant optimality claim does not hold.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The closed-form gains for focusing, steering and con-focusing, the crossing point at 1.947, the 10 dB/decade scaling, and the claim that con-focusing attains the rank-one eigenbound in leading constant are all derived under the quadratic (Fresnel) phase approximation to the exact distance. This approximation error scales with the fourth power of aperture size over distance and therefore grows with the Fresnel number itself. The paper provides no explicit error bound or higher-order remainder term, so the asymptotic comparison and optimality statement rest on an unquantified regime of validity.","agreement_with_reader":"partial"},"referee_report":{"model":"grok-4.3","summary":"The paper analyzes power transfer in single-user LOS MIMO links with extended apertures at both ends under fully analog unit-modulus beamforming in the Fresnel regime. It derives closed-form power gains for conventional focusing (phase-matched to the other's center) and far-field steering (planar phase ramp), showing these are governed solely by the link Fresnel number and aperture ratio. For equal apertures the gains cross once at the universal value 1.947; beyond this point focusing loses 10 dB per decade. The paper introduces con-focusing (both apertures aimed at the common point subtending equal angles), claiming it attains the rank-one eigenbound in the leading constant, requires no channel knowledge, degenerates to steering for equal apertures, and is obtainable in one beam-refinement round.","tokens_in":1895,"tokens_out":434,"duration_ms":32882,"significance":"If the closed-form derivations and proofs are valid, the work is significant for overturning the MISO-derived intuition that focusing is always preferable in the near field when both terminals have extended apertures. The reduction to two dimensionless parameters, the explicit crossover value, the 10 dB/decade scaling, and the parameter-free con-focusing construction that matches the eigenbound leading term are all strengths. The result supplies a concrete, implementable alternative to focusing with clear regime boundaries.","major_comments":[{"comment":"Abstract and the Fresnel-regime derivations: the closed-form gains, the exact crossing at 1.947, the 10 dB/decade scaling, and the claim that con-focusing attains the rank-one eigenbound in the leading constant are all obtained under the quadratic (Fresnel) phase approximation to the exact distance. The approximation error scales with the fourth power of aperture size over distance and therefore grows with the Fresnel number itself. No explicit remainder bound or higher-order term is supplied, so the asymptotic comparison and optimality statements rest on an unquantified regime of validity.","section":"Abstract and Fresnel-regime analysis"}],"minor_comments":[],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the thoughtful review and for highlighting the reliance on the Fresnel approximation. We address the single major comment below.","responses":[{"response":"We agree that all closed-form expressions and the stated comparisons (including the crossing value, scaling, and leading-constant optimality of con-focusing) are derived under the standard quadratic-phase Fresnel approximation to the exact Euclidean distance. The phase error is indeed O((D/z)^4) where D denotes aperture size and z the link distance, so the relative error grows with the Fresnel number. The manuscript defines its scope as the Fresnel regime, in which the quadratic term dominates the phase; however, we did not supply an explicit remainder bound or a quantitative statement of the Fresnel-number range over which the leading-term asymptotics remain accurate to a prescribed tolerance. In the revised version we will add a short appendix deriving the next-order term in the distance expansion and stating the condition under which the omitted term is negligible relative to the retained quadratic term for the power-gain expressions.","revision_made":"yes","referee_comment":"[Abstract and Fresnel-regime analysis] Abstract and the Fresnel-regime derivations: the closed-form gains, the exact crossing at 1.947, the 10 dB/decade scaling, and the claim that con-focusing attains the rank-one eigenbound in the leading constant are all obtained under the quadratic (Fresnel) phase approximation to the exact distance. The approximation error scales with the fourth power of aperture size over distance and therefore grows with the Fresnel number itself. No explicit remainder bound or higher-order term is supplied, so the asymptotic comparison and optimality statements rest on an unquantified regime of validity."}],"tokens_in":1502,"tokens_out":374,"duration_ms":23454,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The key point is that in single-user LoS MIMO with extended apertures on both sides, focusing loses to steering once the Fresnel number passes 1.947 for equal apertures, and the paper gives a new con-focusing method that works better across the range.\n\nThe work derives closed-form power gains for focusing and steering under the Fresnel approximation, showing the comparison depends only on the Fresnel number and aperture ratio, not on the number of elements. For equal apertures the crossover is exact at 1.947, and beyond that focusing drops 10 dB per decade. Con-focusing, where both point to the common angle point, reaches the leading constant of the rank-one eigenbound, needs no channel knowledge, and falls back to steering when apertures match. It can be found with one beam refinement step without exchanging geometry info. These are concrete results that stand on their own.\n\nThe math appears reproducible from the dimensionless setup, and the paper avoids fitting parameters. The practical angle on acquisition is a plus.\n\nThe main soft spot is the reliance on the quadratic phase approximation. The error term grows with the fourth power of the aperture sizes relative to distance, so it gets worse exactly where the Fresnel number is large and the comparisons matter most. Without an explicit bound or comparison to the exact distance expression, the optimality claims for large Fresnel numbers rest on an unquantified assumption. If the full paper has a validation against the exact model, that would fix it; otherwise it's a gap.\n\nThis paper is aimed at people working on near-field beamforming in wireless systems, especially those dealing with large arrays at both link ends. A reader who wants explicit gain formulas and a new strategy that is simple to implement will find it useful. It has enough new math and a clear practical hook to merit sending out for peer review, though the approximation issue should be addressed in revision.","headline":"Paper introduces con-focusing for better power transfer in extended near-field MIMO and pins the crossover at 1.947, though the quadratic phase approx may limit the large-Fresnel claims.","tokens_in":2354,"tokens_out":451,"would_cite":false,"duration_ms":26637,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"In line-of-sight near-field MIMO with extended apertures on both sides, con-focusing both arrays at their equal-angle point outperforms traditional focusing and reaches the optimal rank-one power gain.","keywords":["near-field MIMO","beamfocusing","line-of-sight channel","analog beamforming","Fresnel number","power gain","con-focusing"],"falsifier":"Measure the power gain achieved by con-focusing at a Fresnel number of 5 with unequal apertures and check whether it stays within a few decibels of the theoretical rank-one upper bound; a larger gap would falsify the order-optimality claim.","tokens_in":2676,"feed_emoji":"📡","tokens_out":533,"duration_ms":36482,"temperature":0.7,"pith_summary":"The paper examines power transfer in a single-user line-of-sight MIMO setup where both the transmitter and receiver have extended apertures. It finds that the standard focusing approach, which phase-matches each array to the center of the other, is beaten by simple far-field steering once the link Fresnel number exceeds a modest threshold. The authors derive exact expressions showing that the crossover depends only on the Fresnel number and the ratio of aperture sizes. They introduce con-focusing, a new phase profile in which both arrays point toward the same intermediate point that makes the subtended angles equal. This method achieves the highest possible leading-term gain for rank-one channels, works without any channel state information, and works with standard beam training procedures.","feed_headline":"Con-focusing outperforms focusing in near-field MIMO","feed_subtitle":"Steering both apertures to an equal-angle point reaches the rank-one power bound without channel knowledge","key_machinery":"Con-focusing, the phase profile in which each aperture is focused toward the axial point from which the two apertures subtend equal angles.","core_discovery":"Under the Fresnel approximation for a line-of-sight MIMO channel with analog unit-modulus beamforming, the received power obtained when both terminals apply the con-focusing phase profile equals the square of the largest singular value of the channel matrix up to a multiplicative constant that does not depend on the Fresnel number. The same profile requires only the knowledge of the link distance and the two aperture lengths and collapses to ordinary plane-wave steering whenever the apertures are identical.","pith_inferences":[],"forward_implications":[],"fun_headline_variants":["Near-field MIMO favors con-focusing over focusing","Con-focusing attains rank-one bound in near-field MIMO","LOS near-field con-focusing reaches eigenbound","Aperture ratio governs focusing vs con-focusing","Con-focusing needs only distance and aperture lengths"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"The analysis assumes a single-user line-of-sight MIMO link free of multipath, fully analog unit-modulus beamforming, and operation in the Fresnel regime.","fun_headline_variants_meta":{"raw":{"variants":["Near-field MIMO favors con-focusing over focusing","Con-focusing attains rank-one bound in near-field MIMO","LOS near-field con-focusing reaches eigenbound","Aperture ratio governs focusing vs con-focusing","Con-focusing needs only distance and aperture lengths"]},"model":"grok-4.3","cost_usd":0.007724,"raw_usage":{"total_tokens":3564,"prompt_tokens":734,"num_sources_used":0,"completion_tokens":72,"cost_in_usd_ticks":77237000,"prompt_tokens_details":{"text_tokens":734,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":2758,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":734,"tokens_out":72,"duration_ms":22243,"temperature":1.0,"reasoning_tokens":2758,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-30T01:36:02.805290+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"Measure the power gain achieved by con-focusing at a Fresnel number of 5 with unequal apertures and check whether it stays within a few decibels of the theoretical rank-one upper bound; a larger gap would falsify the order-optimality claim.","supporting_citations":[],"review_version":1}