期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2017
卷号：5
期号：1
页码：794
DOI：10.15680/IJIRCCE.2017.0501169
出版社：S&S Publications
摘要：Data explosion in current year’s leads in the direction of increasing requirements designed for big dataprocessing in current data centers with the purpose are frequently spread on different geographic regions. The unstabledevelopment of demands on big data processing requires an important burden on computation, storage, andcommunication in information centers, which therefore acquire significant prepared overheads in the direction of datacenter providers. Consequently, cost minimization has developed into a developing issue designed for the approachingbig data era. This big data provider is varied from traditional cloud services, since one of the most importantcharacteristic of big data services is the fixed coupling among data and computation as computation tasks is able to beperformed simply when the related data is presented. Consequently, four factors, i.e., task assignment, databasemanagement, data placement and data movement, extremely manipulate the operational costs of information centers. Inthis paper, develops and introduces a new data processing schema which performed based on the procedure of fuzzytokens which divides and splits the database files into many small number of fuzzy tokens in the direction of handledatabase files in the big database .Each and every one of the fuzzy tokens is denoted as the information of database filesin the big database with three levels such as low, medium and high. The tokens are represented depending on thekeyword given by used in the data placement, task assignment; data center resizing and routing in the direction ofreduce the overall computation cost in huge-scale geo- distributed data centers designed for big data applications. In thedirection of describe the task completion time by means of the consideration of together data transmission andcomputation, here introduces and develops a new two-dimensional Markov chain which derives the average taskcompletion time in closed-form. Additionally, problem is represented and illustrated based on the Mixed-Integer Non-Linear Programming (MINLP). The high efficiency is achieved by using fuzzy tokens in the MINLP and validated bymeans of extensive simulation based studies. The proposed schema fuzzy tokens in the MINLP not only maintain thetask completion time, it moreover manages files in the big database based on the fuzzy tokens in the database.
关键词：big data processing; Data explosion; cost minimization; fuzzy tokens; Mixed-Integer Non-Linear;Programming (MINLP) ; big data applications .
