PDM: Privacy-Aware Deployment of Machine-Learning Applications for Industrial Cyber-Physical Cloud Systems

Xiaolong Xu; Ruichao Mo; Xiaochun Yin; Mohanmmad R. Khosravi; Fahimeh Aghaei; Victor Chang; Guangshun Li

doi:10.1109/TII.2020.3031440

PDM: Privacy-Aware Deployment of Machine-Learning Applications for Industrial Cyber-Physical Cloud Systems

Xiaolong Xu, Ruichao Mo, Xiaochun Yin, Mohanmmad R. Khosravi, Fahimeh Aghaei, Victor Chang, Guangshun Li^*

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

Abstract

The cyber-physical cloud systems (CPCSs) release powerful capability in provisioning the complicated industrial services. Due to the advances of machine learning (ML) in attack detection, a wide range of ML applications are involved in industrial CPCSs. However, how to ensure the implementation efficiency of these applications, and meanwhile avoid the privacy disclosure of the datasets due to data acquisition by different operators, remain challenging for the design of the CPCSs. To fill this gap, in this article a privacy-aware deployment method (PDM), named PDM, is devised for hosting the ML applications in the industrial CPCSs. In PDM, the ML applications are partitioned as multiple computing tasks with certain execution order, like workflows. Specifically, the deployment problem is formulated as a multiobjective problem for improving the implementation performance and resource utility. Then, the most balanced and optimal strategy is selected by leveraging an improved differential evolution technique. Finally, through comprehensive experiments and comparison analysis, PDM is fully evaluated.

Original language	English
Pages (from-to)	5819-5828
Number of pages	10
Journal	IEEE Transactions on Industrial Informatics
Volume	17
Issue number	8
Early online date	15 Oct 2020
DOIs	https://doi.org/10.1109/TII.2020.3031440
Publication status	Published - 1 Aug 2021

Bibliographical note

© 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

Funding Information:
This work was supported in part by the Financial and Science Technology Plan Project of Xinjiang Production and Construction Corps under Grant 2020DB005 in part by the National Natural Science Foundation of China under Grant 61702277, and in part by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD) fund.

Keywords

Cyber-physical cloud systems (CPCSs)
machine learning (ML)
nondominated sorting differential evolution (NSDE)
privacy-aware deployment

Access to Document

10.1109/TII.2020.3031440

Privacy-Aware Deployment of Machine-Learning Applications
© 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
Accepted author manuscript, 5.41 MB

Cite this

@article{164f4f1d65d047e2b9d36091859489f8,

title = "PDM: Privacy-Aware Deployment of Machine-Learning Applications for Industrial Cyber-Physical Cloud Systems",

abstract = "The cyber-physical cloud systems (CPCSs) release powerful capability in provisioning the complicated industrial services. Due to the advances of machine learning (ML) in attack detection, a wide range of ML applications are involved in industrial CPCSs. However, how to ensure the implementation efficiency of these applications, and meanwhile avoid the privacy disclosure of the datasets due to data acquisition by different operators, remain challenging for the design of the CPCSs. To fill this gap, in this article a privacy-aware deployment method (PDM), named PDM, is devised for hosting the ML applications in the industrial CPCSs. In PDM, the ML applications are partitioned as multiple computing tasks with certain execution order, like workflows. Specifically, the deployment problem is formulated as a multiobjective problem for improving the implementation performance and resource utility. Then, the most balanced and optimal strategy is selected by leveraging an improved differential evolution technique. Finally, through comprehensive experiments and comparison analysis, PDM is fully evaluated.",

keywords = "Cyber-physical cloud systems (CPCSs), machine learning (ML), nondominated sorting differential evolution (NSDE), privacy-aware deployment",

author = "Xiaolong Xu and Ruichao Mo and Xiaochun Yin and Khosravi, {Mohanmmad R.} and Fahimeh Aghaei and Victor Chang and Guangshun Li",

note = "{\textcopyright} 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. Funding Information: This work was supported in part by the Financial and Science Technology Plan Project of Xinjiang Production and Construction Corps under Grant 2020DB005 in part by the National Natural Science Foundation of China under Grant 61702277, and in part by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD) fund. ",

year = "2021",

month = aug,

day = "1",

doi = "10.1109/TII.2020.3031440",

language = "English",

volume = "17",

pages = "5819--5828",

journal = "IEEE Transactions on Industrial Informatics",

issn = "1551-3203",

publisher = "IEEE",

number = "8",

}

TY - JOUR

T1 - PDM

T2 - Privacy-Aware Deployment of Machine-Learning Applications for Industrial Cyber-Physical Cloud Systems

AU - Xu, Xiaolong

AU - Mo, Ruichao

AU - Yin, Xiaochun

AU - Khosravi, Mohanmmad R.

AU - Aghaei, Fahimeh

AU - Chang, Victor

AU - Li, Guangshun

N1 - © 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. Funding Information: This work was supported in part by the Financial and Science Technology Plan Project of Xinjiang Production and Construction Corps under Grant 2020DB005 in part by the National Natural Science Foundation of China under Grant 61702277, and in part by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD) fund.

PY - 2021/8/1

Y1 - 2021/8/1

N2 - The cyber-physical cloud systems (CPCSs) release powerful capability in provisioning the complicated industrial services. Due to the advances of machine learning (ML) in attack detection, a wide range of ML applications are involved in industrial CPCSs. However, how to ensure the implementation efficiency of these applications, and meanwhile avoid the privacy disclosure of the datasets due to data acquisition by different operators, remain challenging for the design of the CPCSs. To fill this gap, in this article a privacy-aware deployment method (PDM), named PDM, is devised for hosting the ML applications in the industrial CPCSs. In PDM, the ML applications are partitioned as multiple computing tasks with certain execution order, like workflows. Specifically, the deployment problem is formulated as a multiobjective problem for improving the implementation performance and resource utility. Then, the most balanced and optimal strategy is selected by leveraging an improved differential evolution technique. Finally, through comprehensive experiments and comparison analysis, PDM is fully evaluated.

AB - The cyber-physical cloud systems (CPCSs) release powerful capability in provisioning the complicated industrial services. Due to the advances of machine learning (ML) in attack detection, a wide range of ML applications are involved in industrial CPCSs. However, how to ensure the implementation efficiency of these applications, and meanwhile avoid the privacy disclosure of the datasets due to data acquisition by different operators, remain challenging for the design of the CPCSs. To fill this gap, in this article a privacy-aware deployment method (PDM), named PDM, is devised for hosting the ML applications in the industrial CPCSs. In PDM, the ML applications are partitioned as multiple computing tasks with certain execution order, like workflows. Specifically, the deployment problem is formulated as a multiobjective problem for improving the implementation performance and resource utility. Then, the most balanced and optimal strategy is selected by leveraging an improved differential evolution technique. Finally, through comprehensive experiments and comparison analysis, PDM is fully evaluated.

KW - Cyber-physical cloud systems (CPCSs)

KW - machine learning (ML)

KW - nondominated sorting differential evolution (NSDE)

KW - privacy-aware deployment

UR - http://www.scopus.com/inward/record.url?scp=85105603236&partnerID=8YFLogxK

UR - https://ieeexplore.ieee.org/document/9226095

U2 - 10.1109/TII.2020.3031440

DO - 10.1109/TII.2020.3031440

M3 - Article

AN - SCOPUS:85105603236

SN - 1551-3203

VL - 17

SP - 5819

EP - 5828

JO - IEEE Transactions on Industrial Informatics

JF - IEEE Transactions on Industrial Informatics

IS - 8

ER -

PDM: Privacy-Aware Deployment of Machine-Learning Applications for Industrial Cyber-Physical Cloud Systems

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this