BibTeX:
@article{PAPAGEORGIOU2020232, title = {On 5G network slice modelling: Service-, resource-, or deployment-driven?}, journal = {Computer Communications}, volume = {149}, pages = {232-240}, year = {2020}, issn = {0140-3664}, doi = {https://doi.org/10.1016/j.comcom.2019.10.024}, url = {https://www.sciencedirect.com/science/article/pii/S0140366419306255}, author = {Apostolos Papageorgiou and Adriana Fernández-Fernández and Shuaib Siddiqui and Gino Carrozzo}
Abstract
Network slicing has been one of the hottest topics in standardization recently, as telecom operators are intensively investigating its usage for operating isolated and independently configurable logical networks, in order to ease and promote the network sharing and neutral hosting business. However, extensive deployments of slice management solutions are often impeded by incompatibilities of the used slice data models, which stem from different definitions and modelling approaches for the slicing concept, e.g., some driven by 3GPP standards, others by other standards or proprietary solutions, and so on. Although various studies on slicing have been performed, none of them has focused on slice data modelling across research and standards. Incompatible slice models do not only limit interoperability but they also reduce the efficiency of network slicing systems. This paper lays a foundation towards more efficient and interoperable network slice modelling by methodically investigating, categorizing, and formally describing core slice modelling approaches, including new modelling suggestions. Subsequently, we analyse their advantages and disadvantages and we propose slice model quality metrics, which we use for performing a case study on our testbed.