Scalable Fast Accurate Localization in Single Site MIMO with Small-Scale Dataset Using a Multi-Head Fourier Neural Operator
Bibliographic record
Abstract
Accurate, real-time user localization is critical for 5G/6G applications in V2X (vehicle-to-everything) networks, autonomous systems, and immersive AR/VR, yet remains challenging due to limited labeled/training data. Fingerprinting using Channel State Information (CSI) offers multipath resilience but suffers from sparse ground-truth collection, long training time, non-real-time inference, and resolution dependency (retraining needed for new antenna/subsystem configurations). To address these issues, we propose a mesh invariant Multi-Head Fourier Neural Operator (MH-FNO) with Angle-Delay Profile (ADP) transformation of the CSI data. The mesh invariance property makes MH-FNO scalable and essentially independent of antenna count and subcarrier number. Our approach achieves dramatic training acceleration through FFT-based spectral convolutions, with per-epoch complexity of O(n • hw log(hw)) (where n is the number of ADP training samples and h × w is the spatial grid size of each sample), versus O(n 3 + n 2 • hw) by the Gaussian Process Regression (GPR). Furthermore, the model generalizes to unseen ADP dimensions and enables real-time inference at O(hw log(hw)) per sample, instead of O(n • hw) using GPR. Unlike convolutional neural networks (CNNs) which require large datasets, the proposed MH-FNO is able to achieve a 79% lower RMSE than GPR in the experiments performed, with the use of only 1% training samples.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".