A Comprehensive Evaluation Model for the Matching Degree Between Supply and Demand of the Computing Power Network Services

Li Hongjuan,Zhu Yan,Wen Jiangong,Guo Zhi
DOI: https://doi.org/10.1109/icetci55101.2022.9832168
2022-01-01
Abstract:The computing power network is a new type of information infrastructure. It allocates and schedules computing resources, storage resources, and network resources among clouds, edges and terminals according to business needs. The services of the computing power network should flow from areas with sufficient computing power to those with insufficient computing power. At present, the research of computing power network focuses on the technical solutions, and there is no study on the matching degree between supply and demand of the computing power network services. A comprehensive evaluation model for the matching degree between supply and demand of the computing power network services is proposed, which based on multi-dimensional data. The model is used to evaluate whether the computing power network services in a region are sufficient or not. Firstly, considering with the two dimensions of computing power demand and computing power supply, four aspects including regional scale and industrial development, digital economy development, storage and computing capacity, communication capacity are subdivided. And then 6 demand indicators and 6 supply indicators are designed to build the indicator system of the model. Secondly, the AHP is used to determine the weights of each level of the indicator system, and the calculation method of the index is given. Thirdly, 22 sample cities are selected from 5 regions in China, and their demand rankings are compared with the digital economy cities list in authoritative reports. The conclusion that the order is basically the same, verifies the model reasonable and valid. Finally, the model is used to analyze the matching degree between supply and demand of the computing power network services in 22 sample cities, which provides a reference for the construction of local infrastructure of computing power network.
What problem does this paper attempt to address?