arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2407.08462cs.LGcs.NI
收藏

基于分布式深度强化学习梯度量化的联邦学习车辆边缘计算

Distributed Deep Reinforcement Learning Based Gradient Quantization for Federated Learning Enabled Vehicle Edge Computing

School of Internet of Things Engineering, Wuxi Institute of Technology(物联网工程学院,无锡科技学院);School of Internet of Things Engineering, Jiangnan University(物联网工程学院,江南大学);Department of Electronic Engineering, Beijing National Research Center for Information Science and Technology, Tsinghua University(电子工程系,信息科学与技术国家研究中心,清华大学);Qualcomm(高通);School of Engineering, University of Kent(工程学院,肯特大学);Department of Electrical and Computer Engineering, Hong Kong University of Science and Technology (HKUST)(电气与计算机工程系,香港科技大学)

Cui Zhang, Wenjun Zhang, Qiong Wu, Pingyi Fan, Qiang Fan, Jiangzhou Wang, Khaled B. Letaief

AI总结:

本文针对联邦学习车辆边缘计算中梯度传输延迟大的问题,提出基于分布式深度强化学习的量化级别分配方案,联合优化总训练时间和量化误差,仿真验证了其可行性和有效性。

摘要:

联邦学习(FL)通过共享车辆本地模型的梯度而非本地数据,可以在一定程度上保护车辆边缘计算(VEC)中车辆的隐私。对于车载人工智能(AI)应用而言,车辆本地模型的梯度通常较大,因此传输如此大的梯度会导致较大的每轮延迟。梯度量化已被提出作为一种有效的方法,通过压缩梯度并减少传输梯度所需的比特数(即量化级别)来降低FL使能的VEC中的每轮延迟。量化级别和阈值的选择决定了量化误差,这进一步影响模型精度和训练时间。为此,总训练时间和量化误差(QE)成为FL使能的VEC的两个关键指标。联合优化FL使能的VEC的总训练时间和QE至关重要。然而,时变的信道条件给解决这一问题带来了更多挑战。在本文中,我们提出了一种基于分布式深度强化学习(DRL)的量化级别分配方案,以优化总训练时间和QE方面的长期奖励。大量仿真确定了总训练时间与QE之间的最优加权因子,并证明了所提方案的可行性和有效性。

英文摘要:

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 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, which further affects the model accuracy and training time. To do so, the total training time and quantization error (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 paper, 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.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑