Reliability Assessment Based on Jump Diffusion Model for Cloud Computing under the Influence of Big Data  
Author Yoshinobu Tamura


Co-Author(s) Shigeru Yamada


Abstract At present, the cloud computing under the influence of big data by using such as OpenStack, Hadoop, NoSQL, etc. is known as a next-generation software service paradigm. However, the effective methods of software reliability assessment considering the interaction between the database software and the cloud software have been only few presented. Recently, the cloud computing is structured by using several software, e.g., Hadoop, NoSQL, OpenStack, and Eucalyptus. These are well-known as the bigdata- targeted processing software and the cloud computing software. The method of component-based reliability assessment for the software such as database and cloud is proposed in this paper. Moreover, we propose jump diffusion process modeling as the method of system-wide reliability assessment considering the entire environment of cloud computing. Then, we deeply perform software reliability analysis based on two kinds of data set in terms of the background factors. Also, we analyze the software failure-occurrence time data and the count data of the cumulative number of detected faults by applying the hazard rate model and jump diffusion process one. Moreover, several numerical examples for the actual data are shown in this paper.


Keywords Big data, Cloud computing, Reliability, Hazard rates, Jump diffusion process model
    Article #:  2199
Proceedings of the 21st ISSAT International Conference on Reliability and Quality in Design
August 6-8, 2015 - Philadelphia, Pennsylvia, U.S.A.