Research On Hadoop Based Telecom Operators Massive Data Processing Techonology And Its Applications

With the rapid development of Mobile Internet and enrichment of various new services, user-generated data in processing telecom service has been growing dramatically. Massive data and datastructure changings bring an enormous challenge to the telecom operators in processing, analyzingand managing data. Traditional data warehouse technique can not efficiently store and process theever growing and new types of service data. Hadoop distributed technology provides a promisingmethod to solve these problems, which has been studied in this thesis with theoreatical analysis andexperimental methods.Firstly, this thesis gives a comprehensive introduction of Hadoop based technology, alongwiththe discussion and analysis of some Hadoop key techniques including HDFS, MapReduce and Hive,etc. Then some traditional massive data processing technologies in telecom enterprise have beenanalzed, an improved Hadoop based massive data processing and analysis handling processes isproposed. The massive data processing model and optimization strategy has been presented withtheoretical analysis and experaiental implementation simultaneously. Experimental results show thatthe improved method can significantly reduced massive data processing time, improve efficiency,and reduce operating costs, etc.