{"id":"275cb0b3-c43e-4958-8b68-cb1b27367a5f","arxiv_id":"2501.14552","paper_version":2,"verdict":"UNVERDICTED","confidence":"MODERATE","novelty_score":3.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A thematic survey maps the 6G literature onto the IMT-2030 framework's five dimensions and finds that 69.9% of screened papers address technology enablers.","lead":"This paper is a structured literature review of sixth-generation (6G) cellular network research, organized around the ITU's IMT-2030 framework. It groups roughly 899 prior papers by theme to help standards bodies, researchers, and regulators see where 6G research effort is concentrated.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The quantitative map in Fig. 2 and Table II rests on a single-pass title classification with no coding protocol or inter-rater check, so the reported subarea percentages are not demonstrably stable.","rationale":"The reader correctly identified the citation threshold (>50 citations) as a limitation, but I elevate the absence of a coding protocol and inter-rater check to the primary load-bearing concern. The central quantitative claim—that technology enablers dominate at 69.9% and that orchestration and security are comparatively underexplored—is derived entirely from the authors' assignment of 899 titles to thematic categories. Citation bias affects corpus composition and is explicitly acknowledged in Section II.A; classification subjectivity is unacknowledged and affects every percentage reported. The paper has independent supporting evidence: the corpus is shared in a public repository, the thematic counts are internally consistent (Table II sums to 100%, and Figure 2 sums to the stated category counts), and the narrative synthesis is clearly organized around the IMT-2030 framework. These strengths do not remove the reproducibility problem for the quantitative map. The reader's UNVERDICTED verdict remains appropriate: the review is useful as a structured synthesis, but its headline statistics cannot be trusted as a stable quantitative map until the classification step is validated. I am not recommending rejection; the paper's qualitative synthesis stands on its own, and the concern is about the precision and reproducibility of the numerical claims, not about plagiarism, fraud, or fatal internal inconsistency.","tokens_in":28461,"tokens_out":3603,"duration_ms":33621,"concrete_test":"Download the 899-title corpus from the linked GitHub repository (reference [32]); have two independent coders, blinded to the authors' assignments, classify a random sample of 100 titles into the six categories of Figure 1 and, for technology-enabler titles, into the nine subareas of Table II using a written codebook derived from Section V definitions; compute Cohen's kappa and recalculate the Table II percentages from consensus labels. If kappa is below 0.6 or any subarea share shifts by more than two percentage points, the map is not reproducible without a formal coding protocol.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section II.A describes the corpus of 899 screened titles but never specifies how titles were assigned to the five thematic dimensions in Figure 1 or to the nine research areas in Table II. No codebook, decision rules, dual screening, or inter-rater reliability measure is reported. The paper reports percentages to two decimal places (e.g., antennas and meta-surfaces 20.22%, network orchestration 4.78%, security and privacy 4.46%), which implies a precision that is only as good as the underlying classification. Many papers legitimately span multiple categories—a reconfigurable-intelligent-surface paper could reasonably be placed under Antennas, Transmission, or Network Architecture—yet each title appears to be assigned to exactly one bucket. If even a modest fraction of titles were reassigned by an independent coder, the rank ordering in Table II and the central conclusion that areas such as orchestration and security are 'underexplored' could change. The citation-threshold limitation is acknowledged in the same section, but the classification-reliability issue is not, and it affects every quantitative entry in the map. This is the most load-bearing step for the paper's central claim to be an accurate thematic synthesis.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a semi-systematic thematic synthesis of 6G literature organized around the IMT-2030 framework. It screens 899 titles published between 2019 and 2025, drawn from IEEE Xplore, Scopus, research consortia, and SDO documents, and classifies them into five thematic dimensions: 6G vision, use cases and requirements, technology enablers, vertical applications, and the research frontier. The quantitative centerpiece is Figure 2 and Table II, which report that 69.9% of screened literature falls under technology enablers, with antennas and meta-surfaces (20.22%), network integration (14.49%), AI (13.54%), and spectrum (13.22%) as the largest subareas, while orchestration (4.78%) and security and privacy (4.46%) are smaller. The paper also summarizes IMT-2030 use cases and technical requirements, outlines architectural trends, discusses vertical sectors, and identifies open research problems.","tokens_in":28650,"tokens_out":4109,"duration_ms":36177,"significance":"If the map is accurate, this is a useful reference for 6G stakeholders because it organizes a large body of literature through the IMT-2030 lens and identifies perceived gaps such as orchestration and security. The paper has concrete strengths: the full list of screened publications is available in an open-access repository [32]; the arithmetic in Figure 2 and Table II is internally consistent, with Table II percentages summing to 100; the methodology and its limitations are explicitly discussed in Section II.A; and the use of primary ITU documents (ITU-R M.2160-0, M.2516-0) grounding the framework is appropriate. However, the quantitative map rests on a single-pass subjective classification that has not been validated for reliability, and the corpus construction has at least one visible inconsistency with the stated citation threshold. The contribution is therefore useful but currently conditional on the reproducibility of the classification.","major_comments":[{"comment":"The classification of the 899 screened titles into five thematic dimensions and nine research areas is not reproducible as reported. No coding protocol, codebook, decision rules, dual screening, or inter-rater reliability measure is provided, and the paper does not state how titles spanning multiple categories (e.g., a reconfigurable-intelligent-surface paper could plausibly be placed under Antennas, Transmission, or Network Architecture) were assigned to exactly one bucket. The second-decimal percentages in Table II (e.g., 20.22%, 4.78%, 4.46%) imply a precision that is unsupported by a single-pass title-based assignment. Because the central claim of the paper is an accurate quantitative map of 6G research emphasis, and the conclusion that orchestration and security are 'underexplored' depends on these percentages, the authors should either provide a validated classification with inter-rater agreement and a stated coding protocol, or explicitly downgrade the quantitative claims to approximate rankings rather than precise percentages.","section":"Section II.A and Table II"},{"comment":"The stated inclusion criterion in Section II.A is 'peer-reviewed journals, magazines, and conference proceedings with a minimum citation count of 50,' yet Table I includes [14] as a screened general survey with only 18 citations, marked with an asterisk. If this is a deliberate exception, the rationale is not given; if it is an oversight, the corpus-construction rule is applied inconsistently. Since every percentage in Figure 2 and Table II is computed from this corpus, the inconsistency should be resolved: either exclude [14] from the screened set, or state and justify separate inclusion criteria for low-cited but recent surveys.","section":"Section II.A and Table I"}],"minor_comments":[{"comment":"The table cross-references are incorrect: Section VI.A says 'Table VII summarizes applications in the social sector' but the social-sector table is Table IV, and Section VII says 'Table VI summarizes 6G's research frontier' but the research-frontier table is Table VII.","section":"Section VI.A and Section VII"},{"comment":"There are typos in key displays: Figure 1 labels 'Network Intelligente' (should be 'Intelligence'), and Table II has 'Modern modulation and codign techniques' (should be 'coding').","section":"Figure 1 and Table II"},{"comment":"Table VII contains multiple spelling errors, including 'difsraction' and 'difrraction' for 'diffraction,' 'Matematerials' for 'Metamaterials,' and 'metamateials' for 'metamaterials.'","section":"Table VII"},{"comment":"The text and Figure 3 caption use 'UA Vs' instead of 'UAVs' (e.g., 'LEO satellites, HAPS, and UA Vs' in Section III.A).","section":"Section III and Figure 3"},{"comment":"Reference [80] is incomplete, with placeholder author names ('F. Author and S. Author') and volume/page fields 'XX'; this must be replaced with the actual publication data.","section":"Reference list"},{"comment":"Reference [63] is labeled as a TechRxiv preprint despite the methodology stating that only peer-reviewed journals, magazines, and conference proceedings were screened; if this source is included, the methodology should clarify how preprints were handled.","section":"Reference [63]"}],"recommendation":"major_revision","confidential_remarks":"The paper's main contribution is a reproducible quantitative map, but the reproducibility is not yet demonstrated. The classification-reliability issue in Section II.A is the core barrier; it can be addressed in revision by adding a coding protocol and inter-rater reliability statistics, or by softening the quantitative claims. The citation-threshold inconsistency in Table I is a smaller but visible issue that should be fixed. The open-access dataset and the use of primary ITU documents are genuine strengths that make a revision worthwhile. The self-citations [106] and [113] are peripheral to the survey's conclusions, but they should be declared or minimized to avoid the appearance of promotional inclusion."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a survey, not a research result, and its quantitative map is less solid than the two-decimal percentages suggest because the title-classification step is undocumented. Still, the IMT-2030 framing is useful and the authors ship the underlying corpus, so it deserves a proper review rather than a desk reject.\n\nWhat's actually new: the paper organizes the 6G survey landscape around the five IMT-2030 themes and computes a breakdown of 899 screened titles—69.9% enablers, 19.8% verticals, 8.5% surveys, etc. That distribution is the closest thing to a new result. The paper also provides a readable synthesis of the IMT-2030 use cases, technical requirements, and architecture trends, plus a GitHub repository with the full list of screened publications. The dataset is the best part; it makes the screening claims checkable.\n\nWhere it's soft: the methodology (Section II.A) describes the search and the >50 citation threshold but never explains how titles were placed into the nine research areas in Table II. No codebook, no dual screening, no inter-rater check. The percentages go to two decimals, which implies a precision the method can't support. For example, a RIS paper could plausibly land under Antennas, Transmission, or Network Architecture, and the \"underexplored\" conclusion about orchestration and security could shift if a modest fraction of titles moved. The stress-test note gets this right. Minor issues: Table I includes [14] with 18 citations despite the stated threshold; the claim that no major 2025 surveys exist is asserted without evidence; and the reference list has a placeholder ([80] with author and page 'XX') and a likely duplicate ([56] and [133] are the same survey). None of this sinks the survey, but the editorial sloppiness is visible.\n\nThe central qualitative message—tech enablers dominate the literature, security and orchestration get less attention—matches what I'd expect from reading the field, so the map is probably broadly correct even if the exact numbers aren't stable. The authors acknowledge the citation-threshold limitation, which is honest.\n\nBottom line: this is for readers who want a quick, standards-oriented orientation to 6G research, not for researchers looking for new technical results. With the reference fixes and a short paragraph disclosing the classification limitations, it would be a fine published survey. Send it to referees; it should not be desk-rejected.","headline":"A useful IMT-2030-oriented map of the 6G literature, but the quantitative distribution rests on an undocumented title-classification step and the reference list has unfinished entries.","tokens_in":29160,"tokens_out":3011,"would_cite":false,"duration_ms":25862,"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":"A quantitative map of 6G research, organized by the IMT-2030 framework","keywords":["6G cellular networks","IMT-2030 framework","thematic synthesis","literature review","technology enablers","network orchestration","security and privacy","standardization"],"falsifier":"Re-run the screening on 6G publications from 2023 to 2025 with fewer than 50 citations and compare the theme shares to Figure 2; if the technology-enabler share drops well below 70% or orchestration and security grow substantially, the map is an artifact of the citation threshold. Alternatively, have two independent coders assign the 899 titles to the nine enabler subareas and measure inter-coder agreement on the largest categories.","tokens_in":28223,"feed_emoji":"📶","tokens_out":4565,"duration_ms":38811,"temperature":0.7,"pith_summary":"This paper argues that the sprawling 6G research literature can be organized and quantified through the IMT-2030 framework, the International Telecommunication Union's official vision for sixth-generation mobile networks. It screens 899 publications and finds that 69.9% of well-established 6G research concentrates on technology enablers, led by antennas and meta-surfaces, network integration, AI, and spectrum, while orchestration and security receive far less attention. The authors offer this thematic map, with an open archive of screened titles, as a structured knowledge base for early standardization discussions among researchers, industry, regulators, and standards bodies. A sympathetic reader would care because the map claims to show where research effort actually sits as 6G standardization begins.","feed_headline":"70% of established 6G research targets tech enablers","feed_subtitle":"A thematic map of 899 papers shows antennas, AI, and spectrum dominate, while orchestration and security lag.","key_machinery":"The machinery is a thematic classification scheme derived from the IMT-2030 framework and applied to a curated corpus. The authors screen 899 titles from IEEE Xplore, Scopus, research consortia, and standards developing organizations, keeping peer-reviewed journal, magazine, and conference papers with more than 50 citations plus seminal white papers and standardization documents, and assign each title to one of five themes and nine enabler subareas. The resulting percentage distribution (Figure 2, Table II) is what carries the argument: it turns a qualitative survey into a quantitative map of research emphasis.","core_discovery":"The central claim is that a thematic synthesis of highly cited 6G literature, organized by the five IMT-2030-aligned themes (vision; use cases, requirements, and architecture; enabling technologies; vertical applications; research frontier), provides an accurate and useful measure of the field's emphasis. The quantitative anchor is Figure 2 and Table II: within the 628 technology-enabler publications, antennas and meta-surfaces account for 20.22%, heterogeneous network integration 14.49%, AI-based network intelligence 13.54%, and spectrum 13.22%, whereas distributed orchestration (4.78%) and security and privacy (4.46%) are comparatively thin. The paper concludes that enabler research is strongly represented while vertical applications (19.8%) and the research frontier (0.9%) are smaller, and uses that asymmetry to recommend where future standards-informed research attention is needed.","pith_inferences":["Because the greater-than-50-citations filter systematically excludes very recent work, the 'underexplored' areas such as orchestration and security may be growing faster than the map shows; a time-bounded rerun would test this.","The same thematic lens could be applied to patent filings or 3GPP contributions to compare academic research emphasis with industry and standardization effort.","A natural extension is to weight the screened corpus by recency or venue quality rather than citation count, which would produce a complementary map of emerging rather than established 6G research."],"forward_implications":["Standards bodies can use the map to see which research areas are saturated—antennas, meta-surfaces, network integration, AI, and spectrum—and which are comparatively underexplored, such as orchestration and security.","The small 0.9% share for the research frontier indicates that most established work is aimed at near-term technology enablers rather than long-range challenges beyond 6G.","The paper's synthesis of the six IMT-2030 use cases gives stakeholders a common vocabulary for linking use cases to technical requirements and enabling technologies.","If the map is accurate, funding and standardization effort could be rebalanced toward network orchestration, security and privacy, and vertical applications, which the authors identify as necessary but underweighted."],"supporting_citations":[{"why":"Defines the IMT-2030 framework, its six use cases, and fifteen technical requirements; this is the organizing lens of the entire review.","marker":"[22]"},{"why":"The seminal 6G vision survey by Saad et al. that serves as the baseline and starting point for the thematic screening of later surveys in Table I.","marker":"[15]"},{"why":"Provides the ITU's overview of future technology trends for IMT-2030, which the paper uses as the source for its nine technology-enabler domains.","marker":"[21]"},{"why":"Outlines emerging usage trends for IMT systems toward 2030 and beyond, grounding the vertical-application categories in Section VI.","marker":"[20]"},{"why":"Open-access archive of the screened 6G literature, making the corpus behind the percentage breakdown transparent and replicable.","marker":"[32]"},{"why":"Supplies the semi-systematic review methodology that justifies the selection and synthesis approach for an emerging field.","marker":"[25]"},{"why":"A white paper on the IMT-2030 vision that represents the standards-developing-organization literature included in the screened corpus.","marker":"[24]"}],"fun_headline_variants":["6G research map: enablers dominate, security lags","Antennas and AI lead 6G research, security thin","IMT-2030 mapping reveals 6G research gaps","6G research: antennas 20%, security under 5%","6G enabler research overshadows security and orchestration"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire map rests on treating a citation count above 50 as a valid proxy for 'well-established' 6G research, and on the authors' subjective assignment of titles to themes; if that threshold hides recent high-quality work or the assignments are unreliable, the percentage breakdown could shift.","fun_headline_variants_meta":{"raw":{"variants":["6G research map: enablers dominate, security lags","Antennas and AI lead 6G research, security thin","IMT-2030 mapping reveals 6G research gaps","6G research: antennas 20%, security under 5%","6G enabler research overshadows security and orchestration"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000469,"raw_usage":{"total_tokens":2291,"prompt_tokens":857,"completion_tokens":1434,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":473,"completion_tokens_details":{"reasoning_tokens":1345}},"tokens_in":473,"tokens_out":1434,"duration_ms":11847,"temperature":1.0,"reasoning_tokens":1345,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T15:02:26.738688+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-run the screening on 6G publications from 2023 to 2025 with fewer than 50 citations and compare the theme shares to Figure 2; if the technology-enabler share drops well below 70% or orchestration and security grow substantially, the map is an artifact of the citation threshold. Alternatively, have two independent coders assign the 899 titles to the nine enabler subareas and measure inter-coder agreement on the largest categories.","supporting_citations":[],"review_version":1}