Saved in:
Bibliographic Details
Main Authors: Lin, Yushen, Ding, Zhiguo
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2601.18097
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917223455522816
author Lin, Yushen
Ding, Zhiguo
author_facet Lin, Yushen
Ding, Zhiguo
contents Straggler synchronization is a dominant wall-clock bottleneck in synchronous wireless federated learning (FL). Under non-IID data, however, aggressively sampling only fast clients may significantly slow convergence due to statistical heterogeneity. This paper studies PASS-enabled FL, where a radiating pinching antenna (PA) can be activated at an arbitrary position along a dielectric waveguide to reshape uplink latencies. We consider a joint optimization of PA placement and client participation to minimize the expected time-to-accuracy, coupling the exact expected maximum round latency via order statistics with a heterogeneity-aware convergence factor. We derive first-order optimality conditions that reveal an explicit tail-latency premium in the KKT recursion, quantifying how latency gaps are amplified by maximum-order-statistic synchronization. Under a latency-class structure, we obtain a within-class square-root sampling law and establish a two-class phase transition where slow-class participation collapses under an explicit heterogeneity-threshold condition as the per-round sample size grows. For PA placement, we prove a piecewise envelope-derivative characterization and provide an exact breakpoint-and-root candidate-enumeration procedure. Simulation results verify the theoretical findings and show that PASS enables more eligible participation, yielding higher wall-clock accuracy.
format Preprint
id arxiv_https___arxiv_org_abs_2601_18097
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Tail-Latency-Aware Federated Learning with Pinching Antenna: Latency, Participation, and Placement
Lin, Yushen
Ding, Zhiguo
Information Theory
Straggler synchronization is a dominant wall-clock bottleneck in synchronous wireless federated learning (FL). Under non-IID data, however, aggressively sampling only fast clients may significantly slow convergence due to statistical heterogeneity. This paper studies PASS-enabled FL, where a radiating pinching antenna (PA) can be activated at an arbitrary position along a dielectric waveguide to reshape uplink latencies. We consider a joint optimization of PA placement and client participation to minimize the expected time-to-accuracy, coupling the exact expected maximum round latency via order statistics with a heterogeneity-aware convergence factor. We derive first-order optimality conditions that reveal an explicit tail-latency premium in the KKT recursion, quantifying how latency gaps are amplified by maximum-order-statistic synchronization. Under a latency-class structure, we obtain a within-class square-root sampling law and establish a two-class phase transition where slow-class participation collapses under an explicit heterogeneity-threshold condition as the per-round sample size grows. For PA placement, we prove a piecewise envelope-derivative characterization and provide an exact breakpoint-and-root candidate-enumeration procedure. Simulation results verify the theoretical findings and show that PASS enables more eligible participation, yielding higher wall-clock accuracy.
title Tail-Latency-Aware Federated Learning with Pinching Antenna: Latency, Participation, and Placement
topic Information Theory
url https://arxiv.org/abs/2601.18097