Federated Learning (FL) over wireless networks suffers from significant training latency and degraded convergence, particularly in blocked propagation environments where reliable wireless transmission is challenging. While Reconfigurable Intelligent Surfaces (RIS) can enhance communication reliability, existing studies rarely examine the trade-off between learning convergence and communication delay under modulation-dependent transmission errors.
This paper investigates a wireless FL system under RIS-assisted blocked-link propagation scenarios, focusing on adaptive modulation and sub-channel allocation for convergence-latency aware communication design. By characterizing the effect of symbol errors on uploaded local gradients, we derive a convergence-related upper bound that reveals the impact of Symbol Error Rate (SER) on FL loss decay.
Based on this result, we formulate a joint convergence-latency optimization problem as a Mixed-Integer Nonlinear Programming (MINLP) problem and solve it using a low-complexity hybrid alternating optimization framework. Extensive experiments on MNIST, CIFAR-10, and Speech Commands show that the proposed scheme consistently achieves faster convergence and higher test accuracy than existing adaptive communication schemes, particularly in complex tasks and challenging wireless scenarios.
Blogger's Review: This paper introduces RIS technology to address the convergence and latency issues in wireless federated learning, demonstrating its effectiveness in complex environments. The research offers new insights for improving the reliability and efficiency of wireless communication, holding significant practical and theoretical implications.