Font Size: a A A

Research On Using Cloud Computing On Massive Data Processing In Telecom Business Analysis Suppotr System

Posted on:2012-10-10Degree:MasterType:Thesis
Country:ChinaCandidate:H CengFull Text:PDF
GTID:2248330374991058Subject:Electronics and Communications Engineering
Abstract/Summary:PDF Full Text Request
As a computing platform, Cloud Computing which can provide dynamic resourcepool, is virtual and high available. It has the characteristics of scale, high reliability,commonality and cheap hardware. Telecom business analysis system is facing thechallenge of providing a powerful computing and storage capacity for growing massdata and complex business. The appearance of Cloud Computing will change vastlyservice models and business models of Telecom industy. It will create a new methodto solve the problems caused by the Telecom business, such as mass data can’t besolved effectively, there exist some “data isolated islands” and there are too manybusiness systems.This paper focuses on the status and prospects of cloud computing application inthe field of communications. Reference to the current situation that the currenttelecom business, the amount of data are increasing rapidly and the existingtelecommunication management analysis system has faced the developmentbottlenecks, the paper raises a BASS(business analysis and support system) planbased on cloud platform and gives a lot of emphasize discusses on the processing ofthe system’s mass data.This paper expounds the research of cloud computing and its present applicationsituation, based on this difference between cloud computing and grid computing bycontrast, it concludes the technology advantage of cloud computing and get theanalysis that the opportunities and challenges of telecom operators it will take. ThisPaper presents a design plan for building analysis system of telecom managementbased on the cloud computing platform, and makes a detailed introduction for the keytechnologies of the program.On the basement of analyzing and summarizing thecauses and types of mass data in Mobile Bass, it optimize the existing ETL processingflow of mobile Bass in the light of its shortage and designed SMB-DP based onseparation mechanism. Combining the Map/Reduce task scheduling model of cloudcomputing technology, aiming at deficiencies of scheduling ETL based on greedyalgorithm, the concept of task priority is introduced, according to the characteristicsof mobile BASS. Considering the task run time and task priority and so on, it putforward an improved algorithm referred to as AGB-ETL based on greedy schedulingETL tasks algorithm and achieve an effective way to deal with mass data based on split parallel processing.Reference to the architecture design idea that "BigCloud"cloud computing platform of China mobile, CB-BASS test platform was set up, andthe mobile BASS billing accounting class inventory data was selected as test data.According this project,the algorithm in this paper was proved feasible and efficient.
Keywords/Search Tags:Cloud Computing, Telecom Industry, BASS(business analysis andsupport system), Map/Reduce, Mass Data
PDF Full Text Request
Related items