{"paper":{"title":"Map2APS: A Physically Grounded Benchmark for Direct Angle Power Spectrum Prediction from Urban Geometry","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Map2APS benchmark enables direct prediction of angle power spectra from urban maps with 0.948 cosine similarity on unseen layouts.","cross_cats":[],"primary_cat":"eess.SP","authors_text":"Junxi Huang, Kailong Wang, Nan Cheng, Ruijin Sun, Xiucheng Wang, Zhisheng Yin","submitted_at":"2026-05-14T15:52:20Z","abstract_excerpt":"Angle power spectrum (APS) characterizes the directional distribution of received signal power and is directly relevant to beam management and MIMO processing. While environment-aware learning has been widely studied for radio maps and path loss, direct map-to-APS prediction still lacks a standardized large-scale benchmark. This paper presents Map2APS, a physically grounded benchmark constructed from intelligent ray-tracing (IRT) path-level propagation records. Map2APS covers 51 equal-height urban maps and approximately 2.55 million Tx--Rx samples, with a strict cross-map split for evaluating "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"On the full held-out test set of 249,993 samples, MS-AReg achieves a cosine similarity of 0.948, a peak location error of 1.20°, and an inference latency of 0.101 ms/sample.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That intelligent ray-tracing simulations provide sufficiently accurate ground truth for real-world angle power spectra and that the 51 equal-height urban maps with cross-map splits adequately represent generalization to unseen city layouts.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Map2APS is a new large-scale benchmark with 2.55 million samples from 51 urban maps for predicting angle power spectra from geometry, featuring a cross-map split and MS-AReg baseline with 0.948 cosine similarity.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Map2APS benchmark enables direct prediction of angle power spectra from urban maps with 0.948 cosine similarity on unseen layouts.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"5fe453248f1cd1c60a385ee4de11c19e4b5bc97114d8b5c5f9837c02f47948b7"},"source":{"id":"2605.14989","kind":"arxiv","version":1},"verdict":{"id":"37ac9be2-1002-49a5-874a-65d3a492145a","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T03:16:48.973836Z","strongest_claim":"On the full held-out test set of 249,993 samples, MS-AReg achieves a cosine similarity of 0.948, a peak location error of 1.20°, and an inference latency of 0.101 ms/sample.","one_line_summary":"Map2APS is a new large-scale benchmark with 2.55 million samples from 51 urban maps for predicting angle power spectra from geometry, featuring a cross-map split and MS-AReg baseline with 0.948 cosine similarity.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That intelligent ray-tracing simulations provide sufficiently accurate ground truth for real-world angle power spectra and that the 51 equal-height urban maps with cross-map splits adequately represent generalization to unseen city layouts.","pith_extraction_headline":"Map2APS benchmark enables direct prediction of angle power spectra from urban maps with 0.948 cosine similarity on unseen layouts."},"references":{"count":13,"sample":[{"doi":"10.1109/access.2013.2260813","year":2013,"title":"Millimeter wave mobile communications for 5g cellular: It will work!","work_id":"b3ee76e9-d3ec-4df1-935c-d776e52f4f6c","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2016,"title":"An overview of signal processing techniques for millimeter wave MIMO systems,","work_id":"4dd60ca6-137f-48c4-86d7-5a7911ebb0fb","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"TR 38.901: Study on channel model for frequencies from 0.5 to 100 ghz,","work_id":"5ec02012-f1dd-46ca-b599-554af6b4678f","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2019,"title":"DeepMIMO: A generic deep learning dataset for millimeter wave and massive MIMO applications,","work_id":"c14a0a1d-4885-4819-96c0-27f510ee9039","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1109/twc.2021.3054977","year":2021,"title":"Radiounet: Fast radio map estimation with convolutional neural networks,","work_id":"88a87cb0-8922-4dff-b8d2-3da757d61bf0","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":13,"snapshot_sha256":"5b91a5266beaa206a507a3e4109677132dd079ce246494443a9f936f14ed4bca","internal_anchors":1},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}