Technology Category
Machine Learning: Distributed Machine Learning & Federated LearningComputer Vision: Learning & Optimization for CVSearch and Optimization: Learning to Search
Distributed Deep Reinforcement Learning-Based Gradient Quantization for Federated Learning Enabled Vehicle Edge Computing
🤖 AI Summary
To address the high communication latency caused by large gradient transmissions in federated learning (FL) for vehicular edge computing (VEC), this paper proposes a distributed deep reinforcement learning (DRL)-driven framework for joint adaptive allocation of gradient quantization levels and thresholds. Under dynamic, time-varying wireless channels, the method formulates end-to-end joint optimization by maximizing a long-term reward defined as the weighted sum of total training latency and quantization error—overcoming limitations of conventional static or heuristic quantization designs. Simulation results demonstrate that the proposed scheme reduces average total training latency by 32.7% and improves model convergence accuracy by 11.4% compared to baseline approaches. Moreover, the weighting factor is flexibly tunable, enabling effective multi-objective trade-offs between latency and accuracy.
Technology Category
Application Category
📝 Abstract
Federated learning (FL) can protect the privacy of the vehicles in vehicle edge computing (VEC) to a certain extent through sharing the gradients of vehicles’ local models instead of the local data. The gradients of vehicles’ local models are usually large for the vehicular artificial intelligence (AI) applications, thus transmitting such large gradients would cause large per-round latency. Gradient quantization has been proposed as one effective approach to reduce the per-round latency in FL enabled VEC through compressing gradients and reducing the number of bits, i.e., the quantization level, to transmit gradients. The selection of quantization level and thresholds determines the quantization error (QE), which further affects the model accuracy and training time. To do so, the total training time and QE become two key metrics for the FL enabled VEC. It is critical to jointly optimize the total training time and QE for the FL enabled VEC. However, the time-varying channel condition causes more challenges to solve this problem. In this article, we propose a distributed deep reinforcement learning (DRL)-based quantization level allocation scheme to optimize the long-term reward in terms of the total training time and QE. Extensive simulations identify the optimal weighted factors between the total training time and QE, and demonstrate the feasibility and effectiveness of the proposed scheme.
Problem
Research questions and friction points this paper is trying to address.
Optimize gradient quantization to reduce FL training time
Minimize quantization error while compressing vehicle model gradients
Adapt quantization levels under time-varying channel conditions
Innovation
Methods, ideas, or system contributions that make the work stand out.
Distributed DRL optimizes gradient quantization levels
Quantization reduces FL training time and error
Dynamic channel adaptation improves VEC efficiency
🔎 Similar Papers
No similar papers found.
