Читать книгу Machine Learning Paradigm for Internet of Things Applications - Группа авторов - Страница 43

The preparation and preparing of information for such interactions is a critical activity. To respond to this issue, various types of data processing, such as edge analytics, stream analysis, and database IoT analysis, must be applied [22]. Computing frameworks play an vital role in connecting the server with neighboring computer structures and frameworks that depend on the location and the processing server where the data is processed. Architecture is basically classified into several categories for the networking and filter data for data centers.

Edge Computing: This approach to computation allows data to initially be stored on edge computers. Edge devices cannot be linked continuously to the network, so a backup of the master data/reference data is required for offline processing.

Cloud Computing: This approach and the design has high latency and high load balance, which means that this architecture is not ideal for the processing of IoT data since it can work for other processing at high speeds.

There are several other type of cloud computing services like Iaas, Paas, and Saas. These are equipped with the data transmission via API or several other SDK kit for the user interface.