Hadoop Method And System Design Of Massive Sensor Information Processing Based On Service Platform Of The Internet Of Things

IOT(Internet of Things) service platform provides customers with a variety of intelligentservices through processing sensor information. The sensor information has its special features: bigdata, complex sources and openness.As a result, single node processing methods or independentprocessing system can not meet requirements. So it is essential to research methods of processingmassive sensor information in IOT service platform. This thesis is done in this context.Cloud computing is the key technology of IOT service platform, users can make use ofdistributed sensor information, computing resources and storage resources as demand. On the basisof integrated analysis of IOT services platform and key techniques of cloud computing, system ofprocessing massive sensor information is proposed. Firstly, a technical framework of the system ispresented. Second, its functional modules are detailed designed using distributed and parallelcomputing.The system takes advantage of the Hadoop, including MapReduce programming model,HBase distributed database and distributed file system HDFS, besides, Oracle database is also usedas a supplement, so that the system achieves high availability and scalability. Finally, the system isimplemented in the intelligent transportation services. Programming and demonstrations shows thatthe system is suitable to massive sensor information processing.At last, conclusion and future research direction is presented.