Cargando…
Evolvable design of network-oriented services based on a core/periphery structure
In recent years, many new network-oriented services have emerged, and such services will need to be virtualized in the multi-access edge computing environment, which is currently being standardized along with fifth-generation network technology. The environment surrounding the service functions netw...
Autores principales: | , , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
Nature Publishing Group UK
2023
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10356835/ https://www.ncbi.nlm.nih.gov/pubmed/37468481 http://dx.doi.org/10.1038/s41598-023-38695-5 |
Sumario: | In recent years, many new network-oriented services have emerged, and such services will need to be virtualized in the multi-access edge computing environment, which is currently being standardized along with fifth-generation network technology. The environment surrounding the service functions network changes over time, such as breaking changes of APIs, and these changes impact the services. The service design should be adaptable to user requirements and environmental changes for accommodating a large number of services at low cost. In addition, it is required not only to assume environmental changes when initially designing the service functions network, but also to enable the network to continue to change its structure to adapt to new environmental changes in the future. In this paper, we propose a method to evolve the entire network of service functions based on a core/periphery structure. The advantage of the core/periphery structure is that it helps reduce the costs for maintaining or changing services by dividing the service functions into core and periphery functions. We propose a method to evolve a service functions network based on this core/periphery structure. Our method evolves the structure of the service functions network at low cost by keeping the core and peripheral functions at the appropriate scale. In addition, our proposed method accommodates almost 100% of randomly generated service chains, and holds their length to less than twice the minimum chain length. Our simulation results reveal that the structure of the service functions networks can continue to evolve at a low cost and maintain a high service accommodation ratio. |
---|