{"id":"bf27f3e0-d10a-4a32-b607-1739eb96270c","arxiv_id":"2606.29332","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":6.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"Derives asymptotically tight capacity bounds for MIMO-OWC channels under average-power constraint via nonnegative basis pursuit characterization of image-vector distributions.","lead":"This paper derives new lower and upper bounds on the capacity of multiple-antenna optical wireless channels under a total average power limit by reformulating the problem around minimum-l1-norm inputs. A smart generalist might read it to see how theoretical bounds can guide the design of faster, more reliable light-based communication links in indoor and outdoor settings.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.3","headline":"No significant objection identified","rationale":"The reader's identified weakest assumption is not load-bearing: NN-BP is defined to solve exactly for the minimum-l1 input inducing a given image, and convexity guarantees the global solution. The abstract-only review led to an overly conservative verdict; the logical structure of the capacity reduction and the high-SNR claim contain no evident internal gap once the full derivation is considered.","tokens_in":1633,"tokens_out":310,"duration_ms":41294,"concrete_test":"Extract the explicit high-SNR expansions of the lower and upper bounds from the proof (likely in the sections deriving the nT >= nR and nT < nR cases); numerically evaluate the difference between these two expansions for SNR > 40 dB and confirm that it tends to zero.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is asymptotic tightness of the derived bounds at high SNR. The NN-BP step is not a hidden assumption but the explicit convex program used to obtain the minimal l1-norm cost for each image vector; by construction it returns the minimum under the linear image constraint and nonnegativity. Replacing any feasible input by its NN-BP solution preserves the output distribution while weakly decreasing average power, so the capacity reduction to an optimization over image distributions is exact. The subsequent bound derivations and high-SNR analysis therefore rest on standard arguments (mutual-information bounds under the effective cost function) rather than on an unverified identification step.","agreement_with_reader":"disagree"},"referee_report":{"model":"grok-4.3","summary":"The paper claims to derive an equivalent capacity expression for MIMO-OWC channels under average power constraint by solving a nonnegative basis pursuit problem for each image vector to find the min-l1-norm input. It then provides computable lower and upper bounds for the cases when the number of transmitters is at least or less than the number of receivers, and proves these bounds are asymptotically tight at high SNR. Numerical results are provided for indoor and outdoor scenarios showing improvement over prior bounds.","tokens_in":1764,"tokens_out":302,"duration_ms":43678,"significance":"If the asymptotic tightness holds, this work offers a practical way to characterize the high-SNR capacity of MIMO optical wireless channels, closing the constant gap that previous bounds left. The NN-BP reformulation is a key contribution as it exactly reduces the problem to optimization over image distributions without sacrificing optimality, enabling the subsequent bound derivations.","major_comments":[],"minor_comments":[{"comment":"The claim that the bounds 'close the constant gap in the high-SNR regime' would be strengthened by a brief mention of the gap size or the previous bounds' behavior.","section":"Abstract"},{"comment":"Notation for nT and nR should be defined at first use in the introduction.","section":null},{"comment":"It would be useful to include the specific SNR values or ranges used in the simulations for reproducibility.","section":"Numerical results"}],"recommendation":"minor_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the positive assessment of the manuscript, the recognition of the NN-BP reformulation as a key contribution, and the recommendation for minor revision. No major comments were listed in the report.","responses":[],"tokens_in":1146,"tokens_out":52,"duration_ms":13065,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"This paper tightens capacity bounds for MIMO optical wireless channels under average power constraint. The core step is the nonnegative basis pursuit reformulation that reduces the problem to optimization over image-vector distributions, then produces explicit lower and upper bounds that hold for both nT >= nR and nT < nR. The bounds are shown to be asymptotically tight at high SNR, and numerical checks in indoor and outdoor settings improve on prior results while closing the constant gap.\n\nThe NN-BP characterization is the part that stands out as new. Because any feasible input can be swapped for its minimum-l1-norm counterpart without altering the output distribution or raising average power, the reduction to image distributions is exact. From there the bound derivations and high-SNR analysis rest on standard mutual-information techniques applied to the effective cost. The stress-test note confirms this step does not introduce hidden assumptions.\n\nSoft spots are limited. The abstract only outlines the derivations, so the full paper must be checked for the concrete constructions of the bounds themselves; if those steps contain gaps the tightness claim could weaken. The work stays within the average-power constraint, which is stated clearly but narrows the scope compared with peak-power models. No circularity or fitting issues appear.\n\nThe paper is aimed at specialists already working on capacity of optical wireless systems. A reader tracking MIMO-OWC bounds will get usable expressions and a cleaner high-SNR picture. It is incremental rather than foundational, but the technical contribution is concrete and the claims are in principle falsifiable.\n\nI would send it to peer review. The reformulation and the explicit bounds for both antenna regimes are solid enough to justify referee time.","headline":"The paper gives tighter computable bounds for MIMO-OWC capacity under average power via an exact NN-BP reformulation and proves high-SNR asymptotic tightness.","tokens_in":2264,"tokens_out":412,"would_cite":false,"duration_ms":33812,"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":"MIMO-OWC capacity equals an equivalent expression over image vectors whose lower and upper bounds become identical at high SNR.","keywords":["MIMO","optical wireless communication","channel capacity","average power constraint","high SNR","nonnegative basis pursuit","capacity bounds"],"falsifier":"A numerical check in which the gap between the proposed lower and upper bounds remains bounded away from zero as SNR is driven to arbitrarily large values would falsify the asymptotic-tightness claim.","tokens_in":2549,"feed_emoji":"📡","tokens_out":658,"duration_ms":34693,"temperature":0.7,"pith_summary":"The paper shows that multiple nonnegative input vectors can produce the same output distribution in MIMO optical wireless channels under average power, so it solves a nonnegative basis pursuit problem to select the minimum-l1-norm input for each image vector. This yields an equivalent capacity formula written solely in terms of the distribution over image vectors. From that formula the authors derive explicit, computable lower and upper bounds that apply separately when the number of transmit antennas is larger or smaller than the number of receive antennas. They then prove the two bounds coincide asymptotically as SNR tends to infinity.","feed_headline":"MIMO-OWC capacity bounds coincide at high SNR","feed_subtitle":"Nonnegative basis pursuit reduces the problem to image-vector distributions and yields asymptotically tight bounds for any antenna ratio.","key_machinery":"The nonnegative basis pursuit (NN-BP) problem, which selects the minimum-l1-norm nonnegative input that produces a given image vector and thereby reduces the original capacity optimization to a search over distributions on image vectors.","core_discovery":"By formulating a nonnegative basis pursuit problem to identify the minimum-l1-norm input vector for each image vector, the channel capacity is equivalently expressed in terms of the image-vector distribution; this permits derivation of computable lower and upper bounds for both nT >= nR and nT < nR cases that are asymptotically tight in the high-SNR regime.","pith_inferences":["The same NN-BP reduction may apply to other nonnegativity-constrained channels whose output distributions are determined by linear images of the input.","If the high-SNR tightness holds, the limiting capacity per dimension is governed by the geometry of the image set rather than the precise noise statistics.","The method supplies a concrete way to test whether capacity-achieving distributions concentrate on a finite number of image vectors."],"forward_implications":["The capacity expression depends only on the distribution over image vectors rather than the full input alphabet.","Explicit, numerically solvable lower and upper bounds exist for both the nT >= nR and nT < nR antenna configurations.","The bounds coincide in the limit of infinite SNR, closing any constant gap that remains at finite SNR.","Numerical evaluation in indoor and outdoor scenarios shows the new bounds improve on prior expressions."],"fun_headline_variants":["MIMO-OWC bounds asymptotically tight at high SNR","NN-BP derives tight MIMO-OWC capacity bounds","Asymptotically tight bounds for MIMO-OWC at high SNR","MIMO-OWC capacity bounds close at high SNR for all antenna counts"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"The nonnegative basis pursuit problem correctly identifies the minimum-l1-norm input vector for every image vector that induces the same output distribution.","fun_headline_variants_meta":{"raw":{"variants":["MIMO-OWC bounds asymptotically tight at high SNR","NN-BP derives tight MIMO-OWC capacity bounds","Asymptotically tight bounds for MIMO-OWC at high SNR","MIMO-OWC capacity bounds close at high SNR for all antenna counts"]},"model":"grok-4.3","cost_usd":0.006451,"raw_usage":{"total_tokens":2987,"prompt_tokens":598,"num_sources_used":0,"completion_tokens":71,"cost_in_usd_ticks":64512000,"prompt_tokens_details":{"text_tokens":598,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":2318,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":598,"tokens_out":71,"duration_ms":25779,"temperature":1.0,"reasoning_tokens":2318,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-30T02:31:51.290897+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"A numerical check in which the gap between the proposed lower and upper bounds remains bounded away from zero as SNR is driven to arbitrarily large values would falsify the asymptotic-tightness claim.","supporting_citations":[],"review_version":1}