{"id":"c635ca5d-5f05-416c-9b90-bf6a21f1646e","arxiv_id":"2501.09759","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"An amplifying and filtering reconfigurable intelligent surface with shared amplifier circuits is fabricated and demonstrated to give more than 20 dB of in-band reflection enhancement and strong out-of-band rejection.","lead":"Researchers built and tested a 4x8 programmable surface that both amplifies in-band wireless signals and filters out-of-band interference while steering reflected beams. The design packs amplification, filtering, and beamforming into one low-cost relay that could extend 6G coverage in signal-shadowed areas.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'over 20 dB energy enhancement' and one-tenth-area miniaturization claims rest on an unvalidated 2 V mimic of a lossy RIS, and the supplement's '>15 dB' conflicts with the main text's '>20 dB'; a true passive baseline measurement is needed.","rationale":"The paper presents a genuinely interesting hardware concept—sub-connected AF-RIS combining amplification, filtering, and 2-bit phase control—and the phase-tuning, filtering, and beam-steering data are broadly self-consistent. The load-bearing issue is specifically the comparison that anchors the headline 'over 20 dB energy enhancement' and the one-tenth-area miniaturization. The experimental baseline used for this comparison is the AF-RIS itself at 2 V bias (Fig. 4e–h), not a passive lossy RIS. The paper provides no measured S-parameters for the 2 V state of the amplifier chain (Fig. S3d does not report the gain at 2 V), so the reader cannot verify that the 2 V reflection amplitude equals the passive lossy RIS's roughly -6 to -4 dB. The internal discrepancy between the main text's '>20 dB' and Supplementary Note S5's '>15 dB' for the same operating condition further undermines the exact enhancement factor. Because the area-reduction claim scales as the square root of power enhancement, a 15 dB rather than 20 dB enhancement would reduce the claimed area advantage from 10× to about 5.6×, materially changing the practical significance. These issues are correctable with a dedicated baseline measurement and careful consistency checking, so conditional acceptance remains appropriate; the proposed hardware itself is not invalidated.","tokens_in":15414,"tokens_out":7882,"duration_ms":69841,"concrete_test":"Fabricate or simulate a true passive lossy-RIS subarray (same slot/patch geometry, with the filtering/amplifying circuit replaced by a direct microstrip connection) and measure its cross-polarized reflection amplitude at 2.8, 3.0, 3.2 GHz for steering angles 0°, 10°, 20°, 30° in the same anechoic setup used for the AF-RIS. Overlay these values on Figs. 4e–h alongside the 2 V and 7 V AF-RIS curves. Then recompute the enhancement at 3.0 GHz as the ratio of the 7 V AF-RIS amplitude to this true passive baseline, and verify whether it is ≥20 dB at all angles. Also resolve the 15 dB vs 20 dB discrepancy by re-reading S5 and the main text to identify whether different baselines were used; if S5 uses the same baseline, one of the two numbers is wrong.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim—'over 20 dB energy enhancement' and the derived one-tenth-area miniaturization—is supported in the main text (pp. 8–9, Figs. 4e–h) by comparing the AF-RIS at 7 V control voltage against a baseline that is not a separately fabricated lossy RIS. The red dashed line is the AF-RIS itself with the control voltage set to 2 V, explicitly to 'mimic' a normal lossy RIS. No measurement shows that the 2 V state reproduces the true passive insertion loss; the only passive reference is the simulated lossy RIS element in Fig. S2d with -6.3 to -3.8 dB amplitude, while the standalone circuit measurement in Fig. S3d shows the cascaded amplifier gain spans -1.1 to 26.5 dB for 1–7 V, and the gain at 2 V is not reported. If the 2 V state has positive gain or a different frequency response, the reported 20 dB gap is not a valid measure of improvement over a passive lossy RIS. This concern is reinforced by an internal inconsistency: the main text claims 'more than 20 dB energy enhancement' (Fig. 4 and Discussion), but Supplementary Note S5 states 'more than 15 dB energy enhancement' for the same 7 V condition. Since the one-tenth-area claim uses the square-law area scaling (Ref. 71), a 5 dB change alters the required area ratio from 10× to about 5.6×; the exact baseline therefore directly controls the paper's headline miniaturization advantage.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes an amplifying and filtering reconfigurable intelligent surface (AF-RIS) that combines in-band signal amplification, out-of-band filtering, and 2-bit phase-controlled beam steering. A 4x8 prototype is fabricated and characterized, and a software-defined-radio relay experiment is used to demonstrate communication improvements. The central quantitative claim is that the AF-RIS provides 'over 20 dB energy enhancement' relative to a lossy RIS of the same size, which the authors use to argue for a one-tenth-area miniaturization advantage based on the square-law area scaling of reflected power.","tokens_in":15729,"tokens_out":2786,"duration_ms":28824,"significance":"If the headline enhancement and filtering claims are fully supported, this is a useful hardware demonstration for active RIS research: it integrates amplification, frequency selectivity, and phase control in a single fabricated array, and the sub-connected power-combining/dividing architecture reduces the number of active components relative to per-element amplifier designs. The paper also provides direct experimental evidence of beam steering, frequency selectivity, and communication-quality improvement. However, the strength of the miniaturization and power-consumption conclusions depends on the validity of the 20 dB enhancement baseline, which is currently not established by an independent measurement of a passive lossy RIS.","major_comments":[{"comment":"The measured 'over 20 dB energy enhancement' is referenced against a proxy baseline, not a separately fabricated lossy RIS. The red dashed line in Figs. 4e-h is described as 'the reflection amplitude of the AF-RIS with the control voltage setting as 2V to mimic the amplitude of normal lossy RIS.' No measurement is shown demonstrating that the 2 V state reproduces the amplitude and frequency response of a true passive lossy RIS. Supplementary Fig. S3d reports cascaded-amplifier gain from -1.1 to 26.5 dB for control voltages of 1 to 7 V, but the gain at 2 V is not reported, so the 2 V state could still provide positive gain or a different frequency response. The only passive reference in the paper is the simulated lossy RIS element in Fig. S2d, not a measured baseline. Because the 20 dB number directly feeds the one-tenth-area miniaturization claim, this baseline should be replaced or supplemented with a direct measurement of a passive lossy RIS (or an equivalent passive structure) under the same measurement conditions, with the resulting enhancement figures reported per steering angle.","section":"Results, 'Amplifying and filtering properties'; Fig. 4e-h; Discussion"},{"comment":"There is an internal inconsistency in the headline enhancement value. The main text states that the AF-RIS provides 'more than 20 dB energy enhancement' at 7 V control voltage for steering angles 0°, 10°, 20°, and 30°, while Supplementary Note S5 states 'more than 15 dB energy enhancement' for the same 7 V condition. Since the one-tenth-area miniaturization claim is derived from the square-law area scaling of Ref. 71, a 5 dB difference changes the required area ratio from about 10x to about 5.6x. The authors should reconcile these numbers and report the measured enhancement values (and the baseline used) explicitly.","section":"Supplementary Note S5 vs. main text (Figs. 4e-h, Discussion)"},{"comment":"No measurement uncertainty or repeatability information is provided. The amplitude responses, Q-factor/K20dB values, and SNR curves appear to be single measurements without error bars or repeated trials. Given that the paper emphasizes 'stable' filtering and 'commendable' relay performance, reporting measurement uncertainty or repeated measurements would materially strengthen the experimental claims, especially for the SNR values in Fig. 6f-g and the amplitude tuning range in Figs. 4b and 4e-h.","section":"Experimental verification; Figs. 4 and 6"}],"minor_comments":[{"comment":"The caption of Fig. 2 lists panels (d)-(g) in a way that does not match the in-text references: the text refers to Fig. 2f as 'Active and passive reflection amplitude' and Fig. 2g as 'E-field intensity distributions,' while the caption assigns those labels differently. Please align the caption with the text and panel labels.","section":"Fig. 2 caption and text"},{"comment":"The sentence 'the reflection beam of the AIM array is switched to different directions' contains a typo: 'AIM' should be 'AF-RIS'.","section":"Methods, 'Measurement setup'"},{"comment":"The term 'Wilkson power divider' should be 'Wilkinson power divider'.","section":"Supplementary Note S3"},{"comment":"The use of '4*8' instead of '4×8' and the inconsistent capitalization of 'rectangle coefficient' versus 'rectangular coefficient' should be made uniform.","section":"Throughout"},{"comment":"The measurement is described as taking place in 'a simple anechoic environment'; given that the setup uses absorbers in an indoor environment rather than a certified anechoic chamber, it would be helpful to state the estimated measurement uncertainty or the degree of isolation achieved.","section":"Methods, 'Measurement setup'"}],"recommendation":"major_revision","confidential_remarks":"The paper describes a substantial hardware effort and the core concept is timely, but the headline enhancement number needs a proper passive baseline and reconciliation with the supplementary text. If the authors provide a direct measured baseline and clarify the enhancement values, the paper could become acceptable. I would not recommend rejection because the central architectural contribution—sub-connected amplification with filtering and phase control—does not depend on the disputed 20 dB figure, but the current manuscript overstates the quantitative miniaturization claim."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is worth a serious look. It combines three things that are usually separate in RIS hardware: in-band amplification, out-of-band filtering, and 2-bit phase control. The subarray architecture with one shared amplifier/filter chain feeding eight elements via power-combining/dividing networks is genuinely new, at least relative to the prior work the authors cite (amplifying RIS and filtering RIS as separate designs). The fabricated 4x8 panel demonstrates beam steering, frequency selectivity, and a working QPSK relay experiment. That is real, reproducible engineering evidence, and the paper gives enough design detail (element geometry, bias circuits, measured S-parameters) for someone else to build on it.\n\nThe main soft spot is exactly the one the stress-test note flags. The \"over 20 dB energy enhancement\" is measured against the AF-RIS itself operated at 2 V control voltage, not against a separately fabricated lossy RIS. The text admits this 2 V state is used to mimic a lossy RIS, but nowhere is it shown that this state actually reproduces the passive insertion loss. The only passive baseline in the paper is the simulated lossy RIS element with -6.3 to -3.8 dB reflection, while the standalone circuit measurement shows the cascaded amplifier gain spans -1.1 to 26.5 dB across 1 to 7 V, and the 2 V point is not reported. If the 2 V state has any gain or a different frequency response, the 20 dB gap is not a valid measure of improvement over a passive RIS. The inconsistency between \"more than 20 dB\" in the main text and \"more than 15 dB\" in Supplementary Note S5 for the same 7 V condition makes the exact number even less reliable. Since the one-tenth-area miniaturization claim is derived from that enhancement number, that part is only as strong as the baseline measurement.\n\nOther minor issues: no error bars or repeated measurements, and some figure references in the text don't match the captions. These are fixable with a revision that includes a real passive RIS measurement and a clear statement of measurement uncertainty.\n\nThe central concept is sound and the hardware work is solid. The quantitative headline needs a proper baseline before the paper can be taken at face value. I'd send this to peer review rather than desk-reject, but I'd make the baseline measurement a required condition for acceptance.","headline":"A useful hardware combination of amplification, filtering, and 2-bit beam steering, but the headline 20 dB gain number rests on a mimicking baseline rather than a real passive RIS.","tokens_in":16270,"tokens_out":1179,"would_cite":false,"duration_ms":13540,"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":"A 4x8 reconfigurable intelligent surface combines in-band amplification, out-of-band filtering, and 2-bit beam steering, delivering over 20 dB energy enhancement over a lossy surface of the same size.","keywords":["reconfigurable intelligent surface","amplifying RIS","filtering RIS","2-bit phase control","beam steering","wireless relay","frequency selectivity","power combining network"],"falsifier":"Build an otherwise identical passive surface with the amplifying and filtering circuit replaced by a plain microstrip line and measure its reflected signal strength in the 2.8–3.2 GHz band; if that measured strength differs substantially from the AF-RIS's 2 V curve, the reported 20 dB enhancement over a lossy RIS is not directly established.","tokens_in":15245,"feed_emoji":"📡","tokens_out":11682,"duration_ms":104375,"temperature":0.7,"pith_summary":"The paper sets out to prove that a single reconfigurable surface can act as a compact wireless relay: it receives a signal, amplifies it inside the wanted 2.8–3.2 GHz band, rejects signals outside that band, and steers the reflected beam with 2-bit phase control. The authors build and measure a 4x8 prototype and report more than 20 dB of energy enhancement relative to a lossy surface of the same size, together with sharp out-of-band rejection. The practical payoff, if the device works as claimed, is that a wireless relay could use one-tenth of the surface area and less power than a conventional reflective surface, because one shared amplifier-and-filter circuit serves eight elements through a power-combining and dividing network. A QPSK video-relay experiment supports the claimed beam steering, amplification, and frequency selectivity.","feed_headline":"Reconfigurable surface relay adds 20 dB gain and filters interference","feed_subtitle":"A 4x8 prototype combines amplification, filtering, and 2-bit beam steering for smaller wireless relays.","key_machinery":"The load-bearing mechanism is the sub-connected AF-RIS subarray architecture: eight radiating elements share one filtering and amplifying circuit through a 1-to-8 power-combining network on reception and a tapered 1-to-8 power-dividing network on transmission, while each element carries its own 0°/90° phase shifter and 0°/180° switch for 2-bit phase states. The two orthogonal slot-coupled paths convert the received polarization to a different reradiated polarization, which isolates the amplifier input from its output and prevents oscillation. This single shared circuit is what delivers amplification and filtering at a fraction of the hardware cost of per-element active RIS designs.","core_discovery":"The paper's core claim is that amplification, filtering, and digital phase control can be combined in one RIS without an amplifier on every element. In each 1x8 subarray, eight slot-coupled patches feed a power-combining network, a single amplifier-and-filter chain, a power-dividing network, and eight independently switched phase shifters; orthogonal input and output polarizations keep the amplifier isolated and stable. The measured reflection amplitude in the passband is more than 20 dB above the same surface operated in a low-gain 'lossy' mode, while out-of-band amplitude falls steeply, and four phase states (0°, 90°, 180°, 270°) steer the beam to 0°, 10°, 20°, and 30°. In a software-defined-radio link test, the surface relays a QPSK signal at 3.0 GHz with 27.1 dB received SNR and suppresses 2.65 GHz and 3.35 GHz signals.","pith_inferences":["Editorial inference: the one-tenth-area comparison is only as strong as the low-voltage 'lossy' proxy; if a true passive baseline measures differently, the area advantage changes accordingly.","Editorial inference: because the amplifier boosts in-band noise along with the signal, AF-RIS gives the largest SNR gain in coverage-limited links; in noise-limited or high-interference regimes the gain will saturate.","Editorial inference: the sub-connected sharing principle trades some beamforming degrees of freedom for hardware economy, so scaling to larger apertures will require multiple amplifier modules and care with oscillation.","Editorial inference: a direct head-to-head measurement against a separately fabricated passive RIS, using error-vector magnitude and throughput as well as SNR, would test the relay claim more stringently."],"forward_implications":["A relay built on an AF-RIS can deliver the same received energy as a lossy RIS with roughly one-tenth the surface area, since received power scales with the square of array area and the surface adds more than 20 dB.","Sharing one filtering and amplifying circuit among eight elements reduces the number of amplifiers, filters, and the power budget compared with fully-connected amplifying RIS designs.","The demonstrated out-of-band rejection protects both the relay link and neighboring systems from interference that a conventional amplifying surface would otherwise amplify.","The 2-bit phase states preserve beam steering at 0°, 10°, 20°, and 30° while amplification and filtering are active, so coverage control and interference control do not trade off.","The wireless-link demonstration indicates that a compact AF-RIS can replace a larger passive RIS or a more complex relay in coverage-limited scenarios."],"supporting_citations":[{"why":"Shows a sub-connected active RIS can be much more energy-efficient than a fully-connected one, justifying one shared amplifier per eight elements.","marker":"Ref. 50"},{"why":"Provides the experimental single-element active-RIS validation that this 4x8 array extends to a practical prototype.","marker":"Ref. 53"},{"why":"Supplies the amplified-reflectarray design principles that the slot-coupled patch and microstrip network adapt.","marker":"Refs. 54–56"},{"why":"Prior reflection-enhancement metasurface designs that the AF-RIS extends by adding filtering and shared circuitry.","marker":"Refs. 60–63"},{"why":"Earlier filtering RIS designs with beam steering but passband loss, which the AF-RIS overcomes by adding amplification.","marker":"Refs. 69,70"},{"why":"Documents the out-of-band interference and security problems of conventional RISs that motivate the filtering requirement.","marker":"Refs. 67,68"},{"why":"Path-loss model stating received power scales with the square of RIS area, from which the one-tenth-area miniaturization claim follows.","marker":"Ref. 71"}],"fun_headline_variants":["Amplifying RIS boosts signal 20 dB while filtering interference","Single-chip RIS array amplifies, filters, and steers beams","4x8 metasurface relays with 20 dB gain and built-in filtering","Amplifying and filtering RIS cuts hardware cost in relay","Amplifying-filtering RIS hits 27 dB SNR in relay test"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The headline 20 dB advantage is measured against the same AF-RIS board with its amplifier voltage reduced to 2 V to imitate a lossy surface, not against a separately fabricated passive RIS, so the enhancement and the derived one-tenth-area claim depend on that proxy being representative.","fun_headline_variants_meta":{"raw":{"variants":["Amplifying RIS boosts signal 20 dB while filtering interference","Single-chip RIS array amplifies, filters, and steers beams","4x8 metasurface relays with 20 dB gain and built-in filtering","Amplifying and filtering RIS cuts hardware cost in relay","Amplifying-filtering RIS hits 27 dB SNR in relay test"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000774,"raw_usage":{"total_tokens":3443,"prompt_tokens":983,"completion_tokens":2460,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":599,"completion_tokens_details":{"reasoning_tokens":2369}},"tokens_in":599,"tokens_out":2460,"duration_ms":17636,"temperature":1.0,"reasoning_tokens":2369,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T22:53:17.218796+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Build an otherwise identical passive surface with the amplifying and filtering circuit replaced by a plain microstrip line and measure its reflected signal strength in the 2.8–3.2 GHz band; if that measured strength differs substantially from the AF-RIS's 2 V curve, the reported 20 dB enhancement over a lossy RIS is not directly established.","supporting_citations":[],"review_version":1}