D013 Prime Elementology — Prime Spectral Descriptor Atlas for Synthetic RF

D013 Prime Elementology — Prime Spectral Descriptor Atlas for Synthetic RF Dataset ID: D013Version: 2.0Dataset Type: Synthetic Research DatasetAuthor: Phan Thành TrungORCID: 0009-0000-7520-6781DOI: 10.5281/zenodo.21569013 1. Overview D013 Prime Elementology — Prime Spectral Descriptor Atlas for Synthetic RF is a modular synthetic dataset designed for the structured representation, simulation, and computational analysis of radio-frequency signals. The atlas converts synthetic RF configurations into a large descriptor space combining: RF acquisition and front-end parameters; waveform and carrier properties; modulation and protocol descriptors; propagation and channel conditions; antenna and array configurations; time-domain and frequency-domain statistics; FFT and power spectral density features; spectral peaks and harmonic descriptors; IQ and constellation features; cyclostationary and correlation descriptors; wavelet and cepstral features; information-theoretic and nonlinear-complexity measures; noise, interference, jamming, and anomaly indicators; graph, embedding, and machine-learning fields; experimental Prime Spectral Descriptors; Prime token, harmonic, resonance, and graph representations; Operator Intelligence, XAI, and benchmark variables. D013 is intended as a computational research substrate rather than a collection of physical laboratory measurements. 2. Current Release Scale The packaged D013 v2.0 release contains: 1,568 parameters 8,192 synthetic RF configurations 19 descriptor modules 12,845,056 logical data cells a consolidated preview containing all 1,568 parameters; a complete machine-readable schema; a formatted data dictionary; deterministic generation rules; benchmark and validation documentation; a scalable synthetic-data generator. Each row represents one synthetic RF configuration.Each column represents one descriptor, parameter, label, index, or experimental computational feature. The current release is therefore a compact but operational model of a much larger RF configuration space. 3. Purpose D013 is designed to support research and development in areas such as: RF signal classification; modulation-family recognition; SNR, bandwidth, power, and channel-quality regression; synthetic jamming and spoofing detection; RF anomaly detection; feature-selection experiments; AI and XAI benchmarking; descriptor-family ablation studies; robustness testing under noise, Doppler, multipath, and interference; graph-based representation learning; experimental Prime Spectral Descriptor research; Operator Intelligence pipeline design; reproducible synthetic-data engineering; software, viewer, and database prototyping. The atlas can be used as a controlled environment in which AI systems are trained, compared, explained, and stress-tested before real RF measurements become available. 4. Synthetic Research Boundary D013 does not claim that its records were acquired from spectrum analyzers, software-defined radios, antennas, radar systems, communication laboratories, or other physical RF instruments. All configurations are computationally generated. The Prime-token, Prime-harmonic, Prime-resonance, Prime-graph, and Operator Intelligence fields are experimental computational representations. They are included for simulation, AI/XAI testing, descriptor research, and hypothesis generation. They must not be interpreted as: experimentally validated physical laws; certified RF measurements; regulatory spectrum evidence; electromagnetic compatibility test results; telecommunications compliance data; proof of physical Prime Elementology mechanisms. The atlas should therefore be cited and described as a synthetic research dataset. 5. Descriptor Architecture D013 v2.0 is divided into 19 modules: Module Description M00 Identity and Provenance M01 Scenario and Environment M02 RF Acquisition and Front End M03 Waveform and Carrier M04 Modulation and Protocol M05 Channel and Propagation M06 Antenna and Array M07 Time-Domain Descriptors M08 FFT and Power Spectral Density M09 Spectral Peaks and Harmonics M10 IQ and Constellation M11 Cyclostationary and Correlation M12 Wavelet and Cepstral M13 Information Theory and Complexity M14 Noise, Interference, and Anomaly M15 Graph, Embedding, and Machine Learning M16 Prime Token and Signature M17 Prime Harmonic and Resonance M18 Operator Intelligence, XAI, and Benchmark This modular design allows researchers to use the complete atlas or isolate individual descriptor families. 6. Scalable Configuration Space RF configuration space is not naturally finite. Unlike a fixed nuclear grid, RF systems contain many continuous or discretizable variables, including: carrier frequency; bandwidth; transmit and receive power; sampling rate; phase; timing; modulation; protocol; antenna geometry; channel model; propagation distance; relative speed; Doppler shift; SNR; interference; multipath; weather; mobility; jamming; receiver impairments. The number of possible configurations therefore depends on the selected discretization grid. A representative large-scale combinatorial grid may include: Configuration Axis Number of States Signal classes 12 Modulation and protocol variants 16 RF bands 32 Bandwidth levels 16 SNR levels 8 Channel models 8 Interference states 8 Mobility states 4 Antenna configurations 4 The resulting configuration count is: 12 × 16 × 32 × 16 × 8 × 8 × 8 × 4 × 4 = 805,306,368 synthetic RF configurations When each configuration is represented by the full set of 1,568 D013 parameters: 805,306,368 configurations × 1,568 parameters = 1,262,720,385,024 logical data cells Maximum Expansion Notice Under the representative combinatorial grid defined above, the D013 architecture can be expanded to 805,306,368 synthetic RF configurations, corresponding to 1,262,720,385,024 logical data cells. This figure should be understood as a defined expansion scenario, not as an absolute physical maximum of the RF universe. The true number of possible RF states is effectively unbounded because continuous variables can always be sampled at finer resolution. The value 1,262,720,385,024 therefore represents a documented high-scale atlas design target derived from a specific finite configuration grid. 7. Recommended Release Levels The D013 architecture may be deployed at several scales: Release Level Configurations Parameters Logical Data Cells Demonstration Release 8,192 1,568 12,845,056 Atlas-Compatible Release 435,092 1,568 682,224,256 Full Research Release 1,000,000 1,568 1,568,000,000 Large Research Release 2,000,000 1,568 3,136,000,000 Extended Atlas Release 5,000,000 1,568 7,840,000,000 Combinatorial Maximum Scenario 805,306,368 1,568 1,262,720,385,024 Large releases should be stored in partitioned formats such as Parquet, Arrow, HDF5, Zarr, or distributed object storage rather than as one monolithic CSV file. 8. AI and XAI Research Design D013 supports comparative experiments between several feature families: Conventional RF Layer Modules M01–M15 contain conventional or computationally derived RF, DSP, signal-processing, graph, and machine-learning descriptors. Prime Descriptor Layer Modules M16–M17 contain experimental Prime-based mappings, including: Prime frequency indices; Prime bandwidth indices; Prime power and noise indices; Prime token vectors; Prime signature hashes; Prime-factor overlap; Prime harmonic vectors; Prime resonance vectors; Prime band graphs; Prime transition paths; Prime locking and residual indices. Operator and Explanation Layer Module M18 contains: operator-pipeline identifiers; operator activation traces; Prime-TB Cut activation fields; synthetic decision scores; explanation vectors; feature-stability measures; benchmark labels; robustness variables; validation status. This separation enables controlled ablation studies: Conventional descriptors only Prime descriptors only Conventional + Prime descriptors Conventional + Prime + Operator/XAI descriptors 9. Suggested Benchmark Tasks Recommended tasks include: Classification signal-class classification; modulation recognition; protocol-family classification; RF-band classification; channel-model classification; interference-type classification. Regression SNR estimation; bandwidth estimation; channel-quality prediction; path-loss estimation; Doppler estimation; anomaly-severity estimation. Detection jamming detection; spoofing detection; out-of-distribution detection; low-SNR detection; multipath-severity detection; receiver-impairment detection. Explainability SHAP-style feature ranking; permutation-importance analysis; feature-rank stability; local explanation fidelity; global explanation fidelity; descriptor-family attribution; Prime versus conventional feature comparison. 10. Data Organization A recommended package structure is: D013/ ├── README.md ├── CITATION.cff ├── LICENSE.txt ├── CHANGELOG.md ├── MANIFEST_SHA256.json │ ├── data/ │ ├── modules_csv_gz/ │ ├── preview/ │ ├── parquet/ │ └── database/ │ ├── schema/ │ └── D013_v2_full_schema.json │ ├── dictionary/ │ ├── D013_v2_Data_Dictionary.xlsx │ └── D013_v2_field_dictionary.csv │ ├── generator/ ├── validation/ ├── benchmark/ ├── examples/ └── docs/ For very large releases, the data should be partitioned by signal class, RF band, scenario, or configuration block. 11. Reproducibility A D013 release should preserve: deterministic random seeds; generator version; schema version; module definitions; field-generation rules; configuration-grid definitions; train, validation, and test splits; file checksums; software dependencies; benchmark protocols; known limitations. The combination of schema, generator, manifest, and documentation allows the atlas to be regenerated or expanded without redefining its architecture. 12. Limitations D013 is intentionally synthetic and therefore has important limitations: simulated distributions may not match real RF hardware; synthetic channel models simplify physical environments; protocol behavior may be abstracted; receiver nonlinearities may be incomplete; regulatory constraints are not fully represented; adversarial scenarios may be simplified; correlations introduced by the generator may make some tasks easier than real-world tasks; performance on D013 does not establish performance on measured RF data; Prime descriptors require separate scientific evaluation before any physical interpretation. These limitations are not defects in the dataset's intended role. They define its use as a controlled computational test environment. 13. Publication and DOI Use D013 may be deposited as a dataset on repositories such as Zenodo, provided that the release clearly states: the data are synthetic; no physical measurements are claimed; generation methods are documented; the configuration grid is defined; the dataset version is identified; the author and ORCID are included; the limitations and intended uses are disclosed. A DOI can provide a stable citation target for the dataset, schema, generator, benchmark protocol, and future version history. 14. Citation Suggested citation: Phan Thành Trung. (2026). D013 Prime Elementology — Prime Spectral Descriptor Atlas for Synthetic RF (Version 2.0) [Synthetic research dataset]. Zenodo. DOI: 10.5281/zenodo.21569013. ORCID: 0009-0000-7520-6781. 15. Author Phan Thành TrungIndependent Researcher, VietnamORCID: 0009-0000-7520-6781 16. Final Statement D013 is not intended to claim that every RF state has already been measured or physically verified. Its purpose is to provide a structured computational atlas in which RF signals, spectra, channels, antennas, disturbances, Prime descriptors, AI outputs, and XAI traces can be represented within one extensible architecture. The current dataset is a working-scale implementation. The larger architecture points toward a combinatorial design space of: 805,306,368 Synthetic RF Configurations and: 1,262,720,385,024 Logical Data Cells under the representative maximum-expansion grid defined in this document.