Learning-to-optimize researcher for AI-native 6G, NR-V2X relay selection and FM spectrum coordination
3
papers here
3
reviewed by Evolve
1
first-author
0
corresponding
—
works · OpenAlex
—
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-15) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.
Research profile
Output—
Publication counts, venue mix, and yearly production totals are not available in the current bibliometric snapshot, so output cannot be scored; this axis would measure the volume, venues, and cadence of the researcher's published work.
Impact—
Citation counts, h-index, and downstream reuse are absent from this snapshot, so impact cannot be scored; the axis would measure how often and how widely other studies cite and build on this researcher's papers.
Focus0/100
The three reviewed papers map onto three fields in equal measure — AI-native 6G network architecture, wireless networking with AI for communications, and radio spectrum management — each at one third of the record, which yields a focus score of 0 (evenly spread, no dominant field). Their shared optimization methodology still keeps the portfolio thematically coherent.
Collaboration32/100
Bibliometric linkage for the three reviewed papers identifies 4 distinct coauthors, with Emiliano Pallotti, Giambattista Amati, Simone Angelini, and Pierpaolo Salvo appearing repeatedly. The collaboration score of 32 therefore reflects a compact, stable team producing optimization-oriented wireless research together, rather than a broad cross-institutional network; conclusions rest on only three arXiv records.
Activity—
Indicators of ongoing productivity, such as publication recency and per-year output trends beyond the three reviewed 2026 preprints, are not in the data, so activity cannot be scored; it would measure sustained research tempo over time.
Reproducibility32/100
Across the 3 papers with a reproducibility assessment, the mean score is 32. The NR-V2X GNN relay study rates medium — a credible simulation without baselines or error bars — while the Learning-to-Optimize architecture paper and the FM coordination paper rate low, the former importing results from prior work and the latter resting on a single confidential scenario.
Research summary
Federica Mangiatordi works at the intersection of wireless communications and optimization, with the current record consisting of three arXiv papers from 2026. The unifying thread is decision-making in congested radio environments: classical optimization — MILP formulations and simulated annealing — is wrapped in learned models so it can run in real time. In the NR-V2X study, a GINE graph neural network supervised by MILP solutions claims near-optimal relay selection with roughly 100× speed-up, a credible simulation result that Evolve notes lacks baselines and error bars. As first author of the FM spectrum coordination paper, Mangiatordi shows that Top-K interference graphs combined with clustered simulated annealing match full simulated annealing at about 6% of the runtime for cross-border power control, though evidence rests on one confidential scenario. A third, more architectural paper argues that Learning-to-Optimize should be treated as a missing layer of AI-native 6G networks, illustrated with an NR-V2X GNN case whose results are imported from the authors' earlier work. Field distribution is even across AI-native 6G architecture, wireless networking with AI, and radio spectrum management. The visible collaboration footprint is compact — four distinct coauthors, with Emiliano Pallotti, Giambattista Amati, Simone Angelini, and Pierpaolo Salvo recurring. No citation, venue, or affiliation data is available, so impact and output remain unscored.
Collaboration pattern
All three reviewed papers are 2026 arXiv records with a compact author pool: 4 distinct coauthors in total, with Emiliano Pallotti, Giambattista Amati, Simone Angelini, and Pierpaolo Salvo recurring across the set. Mangiatordi is first author on the FM spectrum coordination paper and a middle author on the two learning-to-optimise studies, none as corresponding author — suggesting a contributing role within a small, stable team rather than independent leadership of the author line.
Research keywords
Highlights
Near-optimal NR-V2X relay selection at ~100× claimed speed-up
First-authored FM cross-border coordination at ~6% of annealing runtime
Position paper: Learning-to-Optimize as a missing layer of AI-native networks
Timeline
- 2026Published an arXiv study on FM spectrum coordination, first-authored: Top-K interference graphs with clustered simulated annealing match full annealing at ~6% runtime.
- 2026Co-authored an arXiv study on real-time relay optimisation in NR-V2X networks via MILP-supervised learning-to-optimise graph neural networks, with a ~100× claimed speed-up.
- 2026Co-authored an arXiv position paper proposing Learning-to-Optimize as the missing architectural layer of AI-native networks, with an NR-V2X GNN case study.
Papers in the Evolve catalog
Frequent co-authors
About this profile
This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.
- All metrics derive from a small bibliometric snapshot covering only the three reviewed 2026 arXiv papers; broader publication records are not included.
- Citation counts, h-index, and venue-level data are unavailable, leaving the output, impact, and activity axes unscored.
- No institution, country, or ORCID was provided, so the profile is name-resolved from public bibliographic data and name ambiguity cannot be excluded.
- Reproducibility scores reflect Evolve's review verdicts on the three papers, not independent replication attempts.