Cloud-fog-edge computing model for video surveillance based on modular arithmetic

Abstract

Information management in many areas, such as smart city, smart manufacturing is closely related to the concept of cloud-fog-edge computing. Video surveillance systems in the concept of a smart city can contain thousands of cameras, and the use of a single data center requires significant computing and network resources to ensure the necessary quality of information. The article considers a distributed computing model in which part of video analytics tasks would be performed in the fog, i.e. on the way to the clouds, on peripheral devices. The use of a redundant residue number system (RRNS) will ensure the reliability of calculations. In the article, the choice of parameters is considered, the modeling of distorted information recovery is carried out, and the prospects of using the residue number system and cloud-fog-edge computing for video analytics tasks are presented