Research Of O-domain Analysis System Based On Big Data Architecture

With the continuous upgrading of mobile communication networks and the promotion of various new intelligent terminals and APPs,the scale of user data carried by operators is expanding,and the data circulated on the network has become an important virtual asset of enterprises.China Mobile needs to integrate data scattered in different systems,complete the deep decoupling of data and applications,and rely on big data to solve current practical problems.The mobile big data platform is based on data processing technologies such as distributed system Hadoop,massively parallel processing system MPP and streaming computing,and integrates data dispersed in various platforms or systems,and accesses them according to unified data model and interface standards,providing flexible services for various business departments and external customers of mobile companies.Service content includes resource service,data service and tool service.By building a unified standardized physical model,the large data platform enlarges and realizes the unified data standard of the whole network.By integrating the signaling data,DPI data,online logs,statistical indicators,user behavior logs and other data of the Operation Support System(OSS)and integrating Business Support Sys,the large data platform can standardize data access.TEM(BSS or B domain)and management support system(MSS or M domain)data enrich the content of data and improve the data source of enterprise-level large data platform.This paper mainly studies the functions of the O-domain analysis system,the related large data platform technology and functional architecture model,as well as the implementation scheme based on large data technology.Firstly,the data analysis and application system in O domain is designed in detail.Then,the related data processing technology of the large data platform is studied and analyzed,and the overall functional architecture is studied and designed according to the relevant functions and technologies.Then,according to the actual situation of a province,the business scale is forecasted,and the hardware model of the big data platform is calculated according to the forecasting results.Finally,an application based on the big data platform in a province is selected for analysis,and the feasibility of the O-domain analysis system based on the big data platform is demonstrated.