A semantic RESTful API for heterogeneous IoT data storage

Internet of health things consists of a sensor device, middleware, cloud storage, and data analytic or IoT applications (apps). Handling distribution, integration, and storing massive heterogeneous data from health sensor devices to cloud storage are some of the challenges in IoT implementation for healthcare. Data from sensor device are possibly structured or unstructured. Our previous work used labeled topic to define the structured or unstructured published data, then stored them in NoSQL database. In this research, a heterogeneous data storage based on the NoSQL database and semantic RESTful API is proposed. A semantic approach is used to define the location of stored data based on topic and types of data. If the data are structured or semi-structured and below 100KB, it will be stored in MongoDB, while the unstructured and over 100KB ones will be stored in GridFS. The test results shown that our proposed system can handle structured data in JSON and unstructured data in JPG. We also conducted a test using run time parameter to measure the performance. Its results concluded that our system worked under the standard based on usability engineering, which is 1000 milliseconds.