Within the scope of the project, we propose to study and build a big data stream processing system for IoT applications deployed on a hybrid environment between cloud computing and fog computing. In addition, we also propose to study techniques to increase the elasticity of resources used for processing big data streams on the hybrid environment in order to increase the system's performance.

Theoretical results: the project has conducted an overview survey on data stream processing methods and systems, resource scaling techniques, and autonomous computing model. These fundamentals are the foundation for building a scalable big data stream processing system for IoT applications in a hybrid cloud and fog environment.
The topic also investigated the deployment contexts of IoT applications that analyze big data in fog computing environment, thereby extracting the specialized functions required in an elastic supporting fog architecture framework. Thereby, the topic introduces AutoFog, an elastic supporting fog architecture framework, with a full range of functional classes that can be flexibly deployed for each specific IoT application case.
Applied results: Based on the results of researching practical big data flow analysis IoT applications such as IoT applications consuming energy in smart homes, the project has implemented the proposed architectural framework with the tier of Data Transport using MQTT and the tier of Data Processing using the Apache Storm platform. The results of experiments show that it is effective in terms of performance when using elasticity techniques to meet the data load demands of these applications.

- Building a system architecture framework to handle big data streams from IoT applications on a hybrid cloud-fog environment that supports resource elasticity.
- Design and implement a test system for the proposed architectural framework for the tier of Data Transport using MQTT broker servers.
- Design and implement a test system for the proposed architecture framework for the tier of Data Processing using the Apache Storm platform.

- The framework processes big data streams from collected IoT dataset on cluster environments: 01

- The framework processes big data streams from collected IoT dataset in the cloud computing environment with the application of resource elasticity techniques: 01

- The framework processes big data streams from collected IoT dataset on a hybrid cloud-fog environment using resource elasticity techniques at both the cloud and fog levels: 01

- Scientific papers in referred international journals: 01

- Scientific papers in referred national journals: 01

- Support for Master's training: 01

- Support for Bachelor's training: 02

The research results of the topic are the premise and basis for further studies on processing big data streams for IoT applications in a cloud-fog hybrid environment using resource scaling techniques. At the same time, these research results can be directly put to use for IoT projects that are being implemented in Vietnam such as smart homes, smart agriculture, or smart health to help enhance processing performance and data analysis of these systems.