Blockchained On-Device Federated Learning

被引：513

作者：

Kim, Hyesung ^{[1
]}

Park, Jihong ^{[2
]}

Bennis, Mehdi ^{[2
]}

Kim, Seong-Lyun ^{[1
]}

机构：

[1] Yonsei Univ, Sch Elect & Elect Engn, Seoul 03722, South Korea

[2] Univ Oulu, Ctr Wireless Commun, Oulu 4500, Finland

来源：

IEEE COMMUNICATIONS LETTERS | 2020年 / 24卷 / 06期

关键词：

Computational modeling; Blockchain; Training; Servers; Nickel; Delays; Data models; On-device machine learning; federated learning; blockchain; latency;

D O I：

10.1109/LCOMM.2019.2921755

中图分类号：

TN [电子技术、通信技术];

学科分类号：

0809 ;

摘要：

By leveraging blockchain, this letter proposes a blockchained federated learning (BlockFL) architecture where local learning model updates are exchanged and verified. This enables on-device machine learning without any centralized training data or coordination by utilizing a consensus mechanism in blockchain. Moreover, we analyze an end-to-end latency model of BlockFL and characterize the optimal block generation rate by considering communication, computation, and consensus delays.

引用

页码：1279 / 1283

页数：5

共 13 条

[1]

[Anonymous], 2016, Federated Optimization: Distributed Machine Learning for On-Device Intelligence

[2] Ultrareliable and Low-Latency Wireless Communication: Tail, Risk, and Scale [J].

Bennis, Mehdi ;

Debbah, Merouane ;

Poor, H. Vincent .

PROCEEDINGS OF THE IEEE, 2018, 106 (10) :1834-1853

[3] On Security Analysis of Proof-of-Elapsed-Time (PoET) [J].

Chen, Lin ;

Xu, Lei ;

Shah, Nolan ;

Gao, Zhimin ;

Lu, Yang ;

Shi, Weidong .

STABILIZATION, SAFETY, AND SECURITY OF DISTRIBUTED SYSTEMS, SSS 2017, 2018, 10616 :282-297

[4]

Debbah M., 2018, ARXIV180708127

[5] Information Propagation in the Bitcoin Network [J].

Decker, Christian ;

Wattenhofert, Roger .

13TH IEEE INTERNATIONAL CONFERENCE ON PEER-TO-PEER COMPUTING (P2P), 2013,

[6]

Luong N.C., 2018, 2018 IEEE International Conference on Communications (ICC), P1

[7]

McMahan B., 2016, PMLR, DOI DOI 10.48550/ARXIV.1602.05629

[8]

McMahan B., 2017, Federated Learning: Collaborative Machine Learning without Centralized Train- ing Data

[9]

Nakamoto S., 2009, BITCOIN

[10]

Park J., 2017, OPT MATER, V5, P1

← 1 2 →