Communication-efficient federated learning for wireless edge and vehicular networks; early-stage bibliometric profile
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
—
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-09-20) 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—
Output is not yet scored because the underlying publication-volume data for this profile is insufficient. This axis would measure the number, cadence, and venue spread of papers; at present only two 2026 arXiv preprints are available for reference.
Impact—
Impact cannot be scored yet because citation-based evidence is not available in this profile. This axis would measure citation counts, h-index-style influence, and uptake of the researcher's federated-learning work by other studies.
Focus0/100
The recorded field distribution splits evenly — 50% machine learning (federated learning and wireless edge systems) and 50% wireless communications and machine learning (communication-efficient federated learning / vehicular edge intelligence) — producing a computed focus score of 0. Both branches still converge on federated learning at the wireless edge, so the low score reflects breadth across adjacent subfields rather than scattered topics.
Collaboration46/100
Collaboration scores 46 based on 9 distinct coauthors recorded in the bibliometric data. Both reviewed papers place Niu in the last author position alongside recurring partners such as Yuxuan Sun, Tan Chen, Sheng Zhou, Yuxuan Bai, and Jintao Yan, indicating a compact, stable working group rather than a wide external collaboration network.
Activity—
Activity is unscored because longitudinal publication data is insufficient. This axis would measure recent output rate and continuity — for example papers per year and gaps between releases — which cannot be computed from the two 2026 preprints currently reviewed.
Reproducibility38/100
Reproducibility scores 38, the mean across 2 reviewed papers. Evolve's reviews cite simulation-only evaluation in both: FedPGT (medium readiness) omits variance reporting and uses narrow sparsification baselines, while Sylvas (low readiness) has a thin, simulation-only study — both clear targets for stronger experimental hygiene.
Research summary
Zhisheng Niu appears in the Evolve catalog as a researcher working at the intersection of federated learning and wireless edge systems, with the profile name-resolved from public bibliographic data; no institution, country, or ORCID is attached. The catalog holds two 2026 arXiv preprints, both with Niu in the last author position. The first, FedPGT, proposes progressive gradient transmission for vehicular federated learning over time-varying channels, combining gradient sparsification, Lyapunov-optimization-based resource allocation, and convergence analysis; Evolve's review credits a coherent theory-to-algorithm pipeline with simulation gains but flags missing variance reporting and narrow sparsification baselines. The second, Sylvas, is a device scheduler for federated continual learning built on drift, collective divergence, and pseudo-label value, judged conceptually sound but thinly evaluated in simulation only. The recorded field distribution splits evenly between machine learning (federated learning and wireless edge systems) and wireless communications and machine learning (communication-efficient federated learning / vehicular edge intelligence), yielding a focus score of 0. The collaboration dimension scores 46 on 9 distinct coauthors, anchored by recurring partners including Yuxuan Sun, Tan Chen, Sheng Zhou, Yuxuan Bai, and Jintao Yan. Reproducibility scores 38, the mean over 2 reviewed papers. Output, impact, and activity lack sufficient data and are left unscored.
Collaboration pattern
Bibliometric records show 9 distinct coauthors and a collaboration score of 46. The frequent-coauthor list forms a compact cluster — Yuxuan Sun, Tan Chen, Sheng Zhou, Yuxuan Bai, and Jintao Yan — and Niu sits in the last author position on both reviewed 2026 papers, a placement often associated with a supervisory role. Current evidence points to a stable in-group pattern rather than broad cross-institution collaboration.
Research keywords
Highlights
Progressive gradient transmission for vehicular federated learning
Synergistic learning value for federated continual learning scheduling
Timeline
- 2026Publishes Sylvas on arXiv: synergistic-learning-value device scheduling for federated continual learning, in the last author position.
- 2026Publishes FedPGT on arXiv: progressive gradient transmission for vehicular federated learning over time-varying channels, in the last author position.
- 2026Evolve reviews both papers, establishing an initial reproducibility baseline of 38 (mean across 2 papers) and coverage in two adjacent fields.
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.
- Bibliometric dimensions are computed from OpenAlex-indexed public bibliographic data only.
- No institution, country, or ORCID is attached; the profile is name-resolved from public bibliographic data.
- Catalog coverage is limited to 2 arXiv preprints from 2026 reviewed by Evolve; output, impact, and activity axes are unscored for lack of data.
- The reproducibility score (38) is a mean over only 2 papers and should be read as a small-sample indicator.