Design And Implementation Of Video Stream System Based On Cloud-Edge Collaboration

With the development of the era of big data of the Internet of Things,the number of devices at the edge of the network has surged,and the amount of data transmitted in the network has also increased.These devices show the characteristics of discrete network distribution and large amount of data,especially the video source devices.Due to the popularity of mobile terminal devices and the deployment of large-scale video source devices,the volume of real-time video data increases exponentially.In the central cloud computing service mode,video data is generated by the video source device at the edge of the network and transmitted to the data center through the network.As the amount of transmitted data increases,the network bandwidth load pressure is huge,the data transmission delay is high,and the real-time video streaming service is poor.The thesis proposes a video streaming system solution based on cloud-edge collaboration to solves the network bandwidth pressure and real-time service problems faced by the central cloud computing in the access scenario of a large number of video source devices.Based on the existing cloud platform technology in the laboratory,combined with the technical characteristics of cloud computing and edge computing,the thesis designs and implements a video streaming system based on cloud-edge collaboration.The cloudedge collaboration solution solves the problems of lack of real-time performance of centralized cloud computing and high network transmission pressure,and also solves the problem that edge computing cannot independently solve massive data tasks due to limited resources.Cloud computing is responsible for non-real-time,high-computing,high-storage video streaming service tasks,and edge computing is responsible for realtime,low-complexity edge discrete video streaming tasks.The terminal video source data is connected to the edge cloud.The edge cloud performs preliminary compression processing on the data,and then synchronizes the central cloud to complete persistent storage.At the same time,the edge supports providing real-time video streaming services.The thesis selects micro-service architecture to design and implement the video streaming system from four aspects.First,cloud-edge communication is the core of cloud-edge collaboration.The thesis designs a unified communication message structure to realize cloud-edge communication collaboration and reduce the coupling between applications in the system.Second,the thesis designs and implements cloud-edge data service collaboration,including video data encoding and packing in edge,cloud-edge data synchronization.Third,the thesis implements an edge video streaming server that supports push streaming,pull streaming and connection security.Finally,following the principle of cloud-side control and edge-side execution,through the cloud-edge communication module to complete command transmission and status reporting,the thesis realizes the cloud-edge management and control collaboration.In addition,the thesis designs detection and alarm for cloud-edge connection anomaly,and uses decentralized database to record the resource status to realize the edge offline autonomy.