【Author】
Cao, Mingrui; Zhang, Long; Cao, Bin
【Source】IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
【Abstract】Due to the distributed characteristics of federated learning (FL), the vulnerability of the global model and the coordination of devices are the main obstacle. As a promising solution of decentralization, scalability, and security, leveraging the blockchain in FL has attracted much attention in recent years. However, the traditional consensus mechanisms designed for blockchain-like proof of work (PoW) would cause extreme resource consumption, which reduces the efficiency of FL greatly, especially when the participating devices are wireless and resource-limited. In order to address device asynchrony and anomaly detection in FL while avoiding the extra resource consumption caused by blockchain, this article introduces a framework for empowering FL using direct acyclic graph (DAG)-based blockchain systematically (DAG-FL). Accordingly, DAG-FL is first introduced from a three-layer architecture in detail, and then, two algorithms DAG-FL Controlling and DAG-FL Updating are designed running on different nodes to elaborate the operation of the DAG-FL consensus mechanism. After that, a Poisson process model is formulated to discuss that how to set deployment parameters to maintain DAG-FL stably in different FL tasks. The extensive simulations and experiments show that DAG-FL can achieve better performance in terms of training efficiency and model accuracy compared with the typical existing on-device FL systems as the benchmarks.
【Keywords】Blockchains; Servers; Data models; Peer-to-peer computing; Wireless networks; Task analysis; Internet; Anomaly detection; asynchrony; blockchain; direct acyclic graph (DAG); federated learning (FL)
【标题】面向设备上的联邦学习：基于直接无环图的区块链方法
【摘要】由于联邦学习（FL）的分布式特性，全局模型的脆弱性和设备的协调性是主要障碍。作为去中心化、可扩展性和安全性的有前途的解决方案，在 FL 中利用区块链近年来备受关注。然而，为类似区块链的工作量证明（PoW）设计的传统共识机制会导致极端的资源消耗，这大大降低了 FL 的效率，尤其是在参与设备是无线和资源有限的情况下。为了解决 FL 中的设备异步和异常检测问题，同时避免区块链带来的额外资源消耗，本文介绍了一个系统地使用基于直接无环图 (DAG) 的区块链 (DAG-FL) 赋能 FL 的框架。因此，首先从三层架构详细介绍 DAG-FL，然后设计了运行在不同节点上的 DAG-FL Controlling 和 DAG-FL Updating 两种算法，详细阐述了 DAG-FL 共识机制的运行。之后，制定了泊松过程模型，讨论如何设置部署参数以在不同的 FL 任务中保持 DAG-FL 的稳定。大量的模拟和实验表明，与作为基准的典型现有设备上 FL 系统相比，DAG-FL 在训练效率和模型精度方面可以取得更好的性能。
【关键词】区块链；服务器；数据模型；对等计算；无线网络;任务分析；互联网;异常检测；异步；区块链；有向无环图（DAG）；联邦学习 (FL)
【文献类型】Article; Early Access
评论