An IoT Middleware of Data Service

With the development of the Internet of Things (IoT), the number of devices and sensors connected to the Internet has increased significantly. In the meantime, the data value and data service are becoming increasingly important in many fields. But the large amount of data in the Internet of Things also brings a lot of issues for diverse data format and different protocols of different devices. So it is necessary to find a solution for shielding the heterogeneity. In this paper, we design and implement a generic middleware, which can connect most devices in IoT due to the "virtual pools" and "protocol adapter". All devices and external systems are abstract into "virtual entities" to eliminate data format heterogeneity. "Entity pool" is used to store, manage and maintain those virtual entities. The middleware provides service invocation for the upper application and supports dynamic perception and discovery of IOT data service, and it designs a mechanism to extract key value from massive data to filter the redundant one. The paper presents experimental results to illustrate the good performance and practicality of the middleware.