The pandora of network slicing: A multicriteria analysis

The evolving requirements of 5G communication business cases around vertical markets in addition to rich content multimedia cloud applications call for advances in enabling technologies and standardization to realize tailor-made end-to-end network services, an approach commonly referred to as network slicing. Different mappings between infrastructure capabilities and custom service requirements produce an overload of information that a network service provider must handle to rationalize possible slicing alternatives for a given business case. This article proposes a methodology exploring algorithms from multicriteria analysis (MCA) to digest a set of slicing criteria advantage points into a coherent set of slicing preferences. Through extensive experimental analysis, we reasoning about observed patterns in the slicing alternatives and discuss the utility of the proposed MCA process for the definition of classes of slices, slicing abstractions views, and recovery/scalability analysis. Altogether, the MCA methodology proposed in this article sheds light into candidate mechanisms to tackle the pandora behind the yet-to-be-understood issues and unanticipated trade-offs of alternative network slicing approaches. Contributions of this article include providing a prominent outlook towards different dimensioning angles of network slicing, scoping its criteria design, timing, and the quality of the criteria information.