For Full-Text PDF, please login, if you are a member of IEICE,|
or go to Pay Per View on menu list, if you are a nonmember of IEICE.
Traffic Pattern Based Data Recovery Scheme for Cyber-Physical Systems
Naushin NOWER Yasuo TAN Azman Osman LIM
IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Publication Date: 2014/09/01
Online ISSN: 1745-1337
Type of Manuscript: PAPER
Category: Systems and Control
data pattern analyzer, stochastic traffic pattern, data recovery, cyber-physical systems,
Full Text: PDF(1.6MB)>>
Feedback data loss can severely degrade overall system performance. In addition, it can affect the control and computation of the Cyber-physical Systems (CPS). CPS hold enormous potential for a wide range of emerging applications that include different data traffic patterns. These data traffic patterns have wide varieties of diversities. To recover various traffic patterns we need to know the nature of their underlying property. In this paper, we propose a data recovery framework for different traffic patterns of CPS, which comprises data pre-processing step. In the proposed framework, we designed a Data Pattern Analyzer to classify the different patterns and built a model based on the pattern as a data pre-processing step. Inside the framework, we propose a data recovery scheme, called Efficient Temporal and Spatial Data Recovery (ETSDR) algorithm to recover the incomplete feedback for CPS to maintain real time control. In this algorithm, we utilize the temporal model based on the traffic pattern and consider the spatial correlation of the nearest neighbor sensors. Numerical results reveal that the proposed ETSDR outperforms both the weighted prediction (WP) and the exponentially weighted moving average (EWMA) algorithms regardless of the increment percentage of missing data in terms of the root mean square error, the mean absolute error, and the integral of absolute error.