Uncertainty Theory Based Partitioning for Cyber-Physical Systems with Uncertain Reliability Analysis
-
Published:2022-05-31
Issue:3
Volume:27
Page:1-19
-
ISSN:1084-4309
-
Container-title:ACM Transactions on Design Automation of Electronic Systems
-
language:en
-
Short-container-title:ACM Trans. Des. Autom. Electron. Syst.
Author:
Chen Si1,
Xie Guoqi1,
Li Renfa1,
Li Keqin2
Affiliation:
1. Hunan University, Changsha, Hunan, China
2. State University of New York, New Paltz, New York, USA
Abstract
Reasonable partitioning is a critical issue for
cyber-physical system (CPS)
design. Traditional CPS partitioning methods run in a determined context and depend on the parameter pre-estimations, but they ignore the uncertainty of parameters and hardly consider reliability. The state-of-the-art work proposed an uncertainty theory based CPS partitioning method, which includes parameter uncertainty and reliability analysis, but it only considers linear uncertainty distributions for variables and ignores the uncertainty of reliability. In this paper, we propose an uncertainty theory based CPS partitioning method with uncertain reliability analysis. We convert the uncertain objective and constraint into determined forms; such conversion methods can be applied to all forms of uncertain variables, not just for linear. By applying uncertain reliability analysis in the uncertainty model, we for the first time include the uncertainty of reliability into the CPS partitioning, where the reliability enhancement algorithm is proposed. We study the performance of the reliability obtained through uncertain reliability analysis, and experimental results show that the system reliability with uncertainty does not change significantly with the growth of task module numbers.
Funder
National Natural Science Foundation of China
Fundamental Research Funds for the Central Universities
Hunan University, China
Publisher
Association for Computing Machinery (ACM)
Subject
Electrical and Electronic Engineering,Computer Graphics and Computer-Aided Design,Computer Science Applications