Latency Optimization for Wireless Federated Learning in Multihop Networks

TL;DR

This paper proposes a joint optimization framework for latency reduction in wireless federated learning over multi-hop networks, incorporating adaptive routing and energy harvesting, achieving up to 69.37% latency savings.

Contribution

It introduces a novel PAFL framework with joint optimization of nodes and routing, including energy harvesting, and develops an efficient algorithm for latency minimization.

Findings

Achieves up to 69.37% latency reduction.

Effective joint optimization of leaf, relay nodes, and routing.

Demonstrates the benefits of adaptive routing and energy harvesting.

Abstract

In this paper, we study a novel latency minimization problem in wireless federated learning (FL) across multi-hop networks. The system comprises multiple routes, each integrating leaf and relay nodes for FL model training. We explore a personalized learning and adaptive aggregation-aware FL (PAFL) framework that effectively addresses data heterogeneity across participating nodes by harmonizing individual and collective learning objectives. We formulate an optimization problem aimed at minimizing system latency through the joint optimization of leaf and relay nodes, as well as relay routing indicator. We also incorporate an additional energy harvesting scheme for the relay nodes to help with their relay tasks. This formulation presents a computationally demanding challenge, and thus we develop a simple yet efficient algorithm based on block coordinate descent and successive convex approximation (SCA) techniques. Simulation results illustrate the efficacy of our proposed joint optimization approach for leaf and relay nodes with relay routing indicator. We observe significant latency savings in the wireless multi-hop PAFL system, with reductions of up to 69.37% compared to schemes optimizing only one node type, traditional greedy algorithm, and scheme without relay routing indicator.