Abstract
Smart video surveillance applications are gaining more attention due to their potential to streamline security and safety operations in public and private spaces. However, these distributed video surveillance systems introduce challenges to network performance, such as data latency, throughput, packet loss and delay. To overcome these challenges, a cloud-edge collaboration architecture can be used to optimize network performance parameters. This architecture separates the control plane from the data plane, enabling distributed streaming of video data between edge nodes and cloud storage nodes. This architecture also features a distributed control plane protocol responsible for forwarding video data between nodes and for network optimization decisions. By using intelligent algorithms, the distributed control plane enables Fine-Grained Intelligence (FGI) that optimizes the network performance parameters based on environment conditions. Such a system can leverage system monitoring and reconfiguration capabilities to optimize network performance in real-time. Furthermore, optimizations include the adjustment of packet size, packet scheduling and route selection, all of which can further reduce latency and throughput. Finally, by using Big Data analytics and Machine Learning algorithms, the cloud-edge collaboration architecture can further adjust the network parameters, resulting in enhanced performance. As a result, this cloud-edge collaboration architecture can be used to optimize network performance parameters and enable smart video surveillance applications.
Keyword
Cloud-Edge collaboration, Cloud Computing, Edge Computing, Artificial Intelligence, Internet of Things.
PDF Download (click here)
|