Software Architecture Analysis And Application Study Based On Fog Computing Platform

Fog computing is a service and application on the edge server between the sensor network and the data center. It has some of the functions of the sensor network and data center and provides limited distributed computing, storing and networking services. As the medium between IoT and Cloud Computing, Fog Computing can solve the problem of the delay for responding the requirement from the end nodes, the heavy burden for storing and computing at cloud server, and the pressure of transmission bandwidth in the networks, and so on. Fog Computing is a new research area. However, there is no unified platform and standard in the academe at present, but there will be a broad application prospects in the big data environment (such as intelligent education, intelligent transportation, smart grid, etc.).This paper mainly focuses on the definition and characteristics of Fog Computing. Starting with the heterogeneous nature of the intelligent terminal, this paper studies the platform architecture and the server construction based on Fog Computing, designs a general architecture model of Fog Computing, and achieves the functionality designed by the Fog Computing through the general network device. Finally, the availability of architecture model for Fog Computing is verified by two implementation examples of data preprocessing and data sharing. The main contributions of this paper are as follows:(1)Design a general architecture model of the Fog Computing, and deploy the Fog Computing at the network transport layer. The purpose of the design is to provide the key functions of Fog Computing such as heterogeneous network conversion services, data pushing services, data storing and computing services. By implementing the proposed general architecture model of Fog Computing, it not only provides the localization intelligent analysis and feedback to the terminal equipment, but also provides data filtering, fusion and other preprocessing services for the cloud servers.(2)According to the Fog Computing model and functions, the paper designed the intelligent gateway with ZigBee, Bluetooth, Wi-Fi and other wireless communication protocols. The intelligent gateway consists of three sink modules and one main module to achieve the networking features of wireless sensor networks, solves data exchanging issues between wireless sensor networks and the Internet, and provides the preprocessing functions of outlier rejection and effective data fusion through combining the algorithm of removing the unusual ones in the sensing data and the algorithm of batch estimation data fusion.The experiment of multi-sensor temperature measurement is done in a room. It demonstrates that the intelligent gateway can deal with sensing the networking access and heterogeneous network switching. In addition, it can accomplish the task of outlier detection and data fusion with high accuracy. It also shows that the Fog Computing could effectively reduce the storing magnitude at the data center, and reduce the networking bandwidth pressure caused by the frequent exchanging of data.(3)Based on the background of Micro-lecture and Flipped Classroom, the Fog Computing model is applied in the intelligent classroom. Gee router is chosen as the fog server to achieve data sharing capabilities. The experiment demonstrates that the fog can effectively reduce the delay caused by data transmission between the cloud and the intelligent terminal, thereby greatly enhance the performance of the entire network, and improve the experience of the actual users.