A successful application of big data storage techniques implemented to criminal investigation for telecom

With the emerging of digital convergence, lots of communication services are generated, and the quantity of data grows rapidly. We face the scalability issue to deal with call data records (CDRs) and so are the other telecom companies. This research uses police CDR query as an example with an intension to increase system execution efficiency and scalability and to reduce total cost by applying cloud service. The implementation applying distributed parallel database (Hive), distributed computing (Hadoop MapReduce), and distributed file system (Hadoop HDFS) will be introduced by a simulation to evaluate the execution efficiency of a query from CDRs. The factors influencing query efficiency, such as the settings of data block size and partition size in HDFS, will also be explored. The experimental results show that applying the big data processing technologies to execute queries from a huge amount of CDRs could improve the system execution efficiency significantly and reduce cost.