The Research on IoT Streaming Data Processing Technology
Mar 14, 2014Author：
The traditional data processing method does not work well in real-time streaming data process in the applications of Internet of Things, but there have no method emerging to solve this challenge now, which would obstruct the large scale application of IoT. This research will focus on theoretical and algorithmic innovation of machine-understandable heterogeneous streaming data processing method for IoT real-time application environment, and verify its availability in the network of intelligent front-ends. A unified semantic description framework of heterogeneous streaming data will be provided, and an access middleware will be developed. The synopsis data structure of the heterogeneous streaming data based on sliding windows model will be constructed, and a stream-density-estimation-based load-adaptive algorithm and evaluation method of real-time performance will be proposed. The layering synopsis data structure and rule matching engine which support multi-user query concurrently will be presented, and rule-based message push service fitted for machine to machine interaction will be designed, an intelligent front-end test bed will be set up. Finally, an IoT-friendly heterogeneous streaming data processing environment will be provided to solve the big challenges of semantic consistency, real-time, and concurrency of cooperative awareness between human and machine and machine and machine.