TY - JOUR
T1 - Incentive Mechanism for Reliable Federated Learning
T2 - A Joint Optimization Approach to Combining Reputation and Contract Theory
AU - Kang, Jiawen
AU - Xiong, Zehui
AU - Niyato, Dusit
AU - Xie, Shengli
AU - Zhang, Junshan
N1 - Funding Information:
Manuscript received June 12, 2019; revised August 15, 2019; accepted September 2, 2019. Date of publication September 11, 2019; date of current version December 11, 2019. This work was supported in part by WASP/NTU under Grant M4082187 (4080), in part by Singapore Ministry of Education (MOE) Tier 1 under Grant 2017-T1-002-007 RG122/17, in part by MOE Tier 2 under Grant MOE2014-T2-2-015 ARC4/15, in part Singapore-Israel NRF-ISF under Grant NRF2015-NRF-ISF001-2277, in part by EMA Energy Resilience under Grant NRF2017EWT-EP003-041, and in part by the Programs of NSFC under Grant 61973087 and Grant 61703113. The work was presented in part at the 16th IEEE Asia Pacific Wireless Communications Symposium 2019. (Corresponding author: Zehui Xiong.) J. Kang, Z. Xiong, and D. Niyato are with the School of Computer Science and Engineering, Nanyang Technological University, Singapore (e-mail: kavinkang@ntu.edu.sg; zxiong002@e.ntu.edu.sg; dniyato@ntu.edu.sg).
Funding Information:
This work was supported in part by WASP/NTU under Grant M4082187 (4080), in part by Singapore Ministry of Education (MOE) Tier 1 under Grant 2017-T1-002-007 RG122/17, in part by MOE Tier 2 under Grant MOE2014-T2-2-015 ARC4/15, in part Singapore- Israel NRF-ISF under Grant NRF2015-NRF-ISF001-2277, in part by EMA Energy Resilience under Grant NRF2017EWT-EP003-041, and in part by the Programs of NSFC under Grant 61973087 and Grant 61703113. The work was presented in part at the 16th IEEE Asia Pacific Wireless Communications Symposium 2019.
Publisher Copyright:
© 2014 IEEE.
PY - 2019/12
Y1 - 2019/12
N2 - Federated learning is an emerging machine learning technique that enables distributed model training using local datasets from large-scale nodes, e.g., mobile devices, but shares only model updates without uploading the raw training data. This technique provides a promising privacy preservation for mobile devices while simultaneously ensuring high learning performance. The majority of existing work has focused on designing advanced learning algorithms with an aim to achieve better learning performance. However, the challenges, such as incentive mechanisms for participating in training and worker (i.e., mobile devices) selection schemes for reliable federated learning, have not been explored yet. These challenges have hindered the widespread adoption of federated learning. To address the above challenges, in this article, we first introduce reputation as the metric to measure the reliability and trustworthiness of the mobile devices. We then design a reputation-based worker selection scheme for reliable federated learning by using a multiweight subjective logic model. We also leverage the blockchain to achieve secure reputation management for workers with nonrepudiation and tamper-resistance properties in a decentralized manner. Moreover, we propose an effective incentive mechanism combining reputation with contract theory to motivate high-reputation mobile devices with high-quality data to participate in model learning. Numerical results clearly indicate that the proposed schemes are efficient for reliable federated learning in terms of significantly improving the learning accuracy.
AB - Federated learning is an emerging machine learning technique that enables distributed model training using local datasets from large-scale nodes, e.g., mobile devices, but shares only model updates without uploading the raw training data. This technique provides a promising privacy preservation for mobile devices while simultaneously ensuring high learning performance. The majority of existing work has focused on designing advanced learning algorithms with an aim to achieve better learning performance. However, the challenges, such as incentive mechanisms for participating in training and worker (i.e., mobile devices) selection schemes for reliable federated learning, have not been explored yet. These challenges have hindered the widespread adoption of federated learning. To address the above challenges, in this article, we first introduce reputation as the metric to measure the reliability and trustworthiness of the mobile devices. We then design a reputation-based worker selection scheme for reliable federated learning by using a multiweight subjective logic model. We also leverage the blockchain to achieve secure reputation management for workers with nonrepudiation and tamper-resistance properties in a decentralized manner. Moreover, we propose an effective incentive mechanism combining reputation with contract theory to motivate high-reputation mobile devices with high-quality data to participate in model learning. Numerical results clearly indicate that the proposed schemes are efficient for reliable federated learning in terms of significantly improving the learning accuracy.
KW - Blockchain
KW - contract theory
KW - federated learning
KW - mobile networks
KW - reputation
KW - security and privacy
UR - http://www.scopus.com/inward/record.url?scp=85076730826&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85076730826&partnerID=8YFLogxK
U2 - 10.1109/JIOT.2019.2940820
DO - 10.1109/JIOT.2019.2940820
M3 - Article
AN - SCOPUS:85076730826
SN - 2327-4662
VL - 6
SP - 10700
EP - 10714
JO - IEEE Internet of Things Journal
JF - IEEE Internet of Things Journal
IS - 6
M1 - 8832210
ER -