Slice Elasticity Architecture (SEA): definition, implementation and integration

Abstract

Network slicing initiatives (e.g., Cloud Network Slicing, 5G Network Slicing) are being widely studied by academia and industry as enablers for future networks like 5G and 6G, as well as verticals such as V2X, critical communications, augmented reality, and others. Several challenges arise with this emerging concept, which needs to be solved to further groundbreaking technologies. A significant challenge is slice elasticity, defined as the capacity to grow or shrink slice resources (e.g., computing, networking, or storage). One alternative to solve this challenge is providing a closed control loop (CCL) involving slice monitoring, management, and orchestration. In addition, this CCL may be easily adapted and coupled with most of the slicing initiatives of standardization bodies. In that sense, this thesis proposes, implements, and integrates the Slice Elasticity Architecture (SEA), an architecture to support a new business model called Slicing Elasticity as a Service (SlEaaS).

This new business model enables slice resource orchestration across a heterogeneous end- to-end infrastructure that spans multiple administrative and technological domains in a

facilitated and automated way for tenants and slice providers. The SEA monitors slice resources periodically, offers the policies/SLAs administration, and performs slice elasticity operations autonomously when the predefined policy is violated to ensure the QoS of the instantiated service. The SEA design is scalable and generic enough to support several technologies transparently. The results demonstrate the SEA functionalities based on scenarios of slice elasticity operations for computing and networking resources.