Intent-Oriented Network Slicing With Hypergraphs

Establishing network slices in line with the intent of particular applications through closed-loop automation represents a key goal on designing future advanced networks. However, the lack of association between application intents and network services imposes challenges to accurately capture user preference and generate customized network slices. This article introduces Intent-Oriented Network Slicing (IONS) based on hypergraph theory to realize the customization of application intents, while maximizing the utilization of network resources. Specifically, IONS introduces a cutting-edge hypergraph framework, utilizing scenarios and patterns as the foundational elements of its hyperedges, which can be further enhanced by creating a dynamic adaptive attention mechanism tailored to the hypergraph’s structure. Through this innovative approach, a sophisticated correlation model is established, linking application intents with network slicing strategies. This model significantly enhances the precision of user acquisition, optimizing network resources and user experience, while thus improving the customization of slice generation. Simulations demonstrate that the generated network slices can accurately reflect application intents.