Space-Earth Integration Network Slice Mapping Based on Multi-Level Attribute Representation Model

• Published in: 2023 8th International Conference on Signal and Image Processing (ICSIP) pp. 905 - 913
• Main Authors: Qu, Hua, Wu, Siying, Zhao, Jihong, Ren, Yayun, Mao, Lin, Wang, Siyan
• Format: Conference Proceeding
• Language: English
• Published: IEEE 08.07.2023
• Subjects: Heuristic algorithms; Image processing; Measurement; Multi-level attributes; Network slicing; Network topology; Probability; slice mapping; Space-Earth Integration Network; Topology
• DOI: 10.1109/ICSIP57908.2023.10271083

Most of the existing network slicing methods only consider a single attribute of network nodes and ignore the overall network attributes, however, due to the different characteristics of network resources, complex node, link and topology states, a wide variety of services, etc., these algorithms may lead to a declining acceptance rate of network slicing and cannot better meet business requirements. To solve these problems, this paper provides a multi-level attribute representation model for Space-Earth Integration Network and a network slicing algorithm based on this model, called M-ARM, which considers multiple attribute characteristics of network slicing nodes. In the first stage, this method uses the resource attributes, topological attributes, location attributes and security attributes of the network slicing nodes to build the multi-level attribute representation model. In the second stage, this model is used to calculate multi-criteria decision metrics and perform node ranking, then the node mapping process in the network slice is performed based on the results of the node rank, and in the third stage, the node location deployment probability is designed and the link mapping process in the network slice is performed using the K-shortest path algorithm to avoid link congestion since frequent shortest path selection. Finally, the results of this algorithm are compared with other algorithms using simulation and simulation experiments. The experiment results show that this algorithm enhances the acceptance rate of network slicing and achieves fine-grained control of the network by the underlying infrastructure resources of the network operator matching the actual service.