Skip to content
DraconiticResearch
Search Sources API
181,686 studies31 sourcesUpdated 14 min ago
PreprintarXiv licencearXiv

Wait to Learn: Timing the Gradient Return in Over-the-Air Split Learning over an Aging Single-Antenna Channel

  • Xianzhong Li
  • Yuzhi Yang
  • Tianhao Guo
Open at sourcePDFDOICite

Abstract

In over-the-air split learning over an aging single-antenna Rayleigh channel, the mean returned gradient equals the antenna-domain gradient scaled by the temporal correlation. Uncorrected training collapses in negative-correlation lobes. Near a correlation zero, no nonzero scalar calibration improves the ratio of mean power to innovation power for an individual transmission, and, on the tested grid, using sixteen times as many pilots recovers only 7 percent of the training gap. Returning at the first correlation extremum with phase correction, 1.3 ms later at 3.5 GHz and 15 m/s, achieves 90.2 percent accuracy, compared with 91.8 percent with CSI. In the tested line-of-sight setting, the near-zero gap becomes negligible.

Related

  1. UAV Fluid-Antenna Channel Acquisition under Intra-Scan Channel Aging
    Yuanhui Wu, Hao Jiang, Liang Wu +1
    arXivPreprintarXiv licenceFull textGeneral mathematicsarXiv
  2. Two-Timescale Antenna Positioning for Pinching-Antenna System Based on Statistical Channel Information
    Jingze Ding, Zijian Zhou, Bingli Jiao +1
    arXivPreprintCC BY 4.0Full textGeneral mathematicsarXiv
  3. Fluid Antenna Secure Channel-driven Constellation (FASCC): a Physical Layer Security approach
    Daniel Dinis, Risto Wichman
    arXivPreprintarXiv licenceFull textGeneral mathematicsarXiv
  4. Fluid Antenna Channel Reconstruction with Non-Negative Least Square Detector
    Yang Chen, Jian Dang, Zaichen Zhang
    arXivPreprintarXiv licenceFull textGeneral mathematicsarXiv
  5. Covert Communications over Enormous Fluid Antenna Systems (E-FAS): Impact of Channel Hardening on Detection
    Farshad Rostami Ghadi, Damoon Shahbaztabar, Kai-Kit Wong +1
    arXivPreprintarXiv licenceFull textGeneral mathematicsarXiv

Cite

BibTeXRIS
BibTeX
@misc{li2026wait,
  title = {{Wait to Learn: Timing the Gradient Return in Over-the-Air Split Learning over an Aging Single-Antenna Channel}},
  author = {Xianzhong Li and Yuzhi Yang and Tianhao Guo},
  howpublished = {arXiv},
  year = {2026},
  month = oct,
  doi = {10.48550/arxiv.2610.23135},
  eprint = {2610.23135},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2610.23135},
}
RIS
TY  - UNPB
TI  - Wait to Learn: Timing the Gradient Return in Over-the-Air Split Learning over an Aging Single-Antenna Channel
AU  - Xianzhong Li
AU  - Yuzhi Yang
AU  - Tianhao Guo
PY  - 2026
DA  - 2026/10/26
PB  - arXiv
DO  - 10.48550/arxiv.2610.23135
UR  - https://arxiv.org/abs/2610.23135
AB  - In over-the-air split learning over an aging single-antenna Rayleigh channel, the mean returned gradient equals the antenna-domain gradient scaled by the temporal correlation. Uncorrected training collapses in negative-correlation lobes. Near a correlation zero, no nonzero scalar calibration improves the ratio of mean power to innovation power for an individual transmission, and, on the tested grid, using sixteen times as many pilots recovers only 7 percent of the training gap. Returning at the first correlation extremum with phase correction, 1.3 ms later at 3.5 GHz and 15 m/s, achieves 90.2 percent accuracy, compared with 91.8 percent with CSI. In the tested line-of-sight setting, the near-zero gap becomes negligible.
ER  -