Hyper-Dynamic End-to-End Slicing - a flexible overlay approach applied to service providers and verticals alike

Network slicing is an end-to-end (E2E) overlay concept that does not only represent one individual network domain but stretches across various ones, such as Fixed, RAN, Core, and Transport, which makes the coordinated realization of intent in a multi-operator and multi-vendor environment a genuine challenge. There will be thousands of slices whose lifecycle management cannot possibly be handled manually, making automation a key enabler of network slicing. This tutorial consists of three parts. First, we will discuss the need for slicing and its automation, both from a business and a technical perspective, the application of slicing to vertical industries, interoperability and openness of slicing (from 3GPP, ETSI ZSM, ONAP, and O-RAN perspectives), as well as existing implementations and security. Secondly, we will elaborate on latest research on hyper-dynamic slicing where the success of a service is not evaluated discretely but as a utility function over requirements such as throughput and latency. We will explain the somewhat counter-intuitive implications of this approach, namely the fact that the optimal partitioning of resources depends on the overall system load and that tenants can differentiate from each other even though they are using the very same physical infrastructure. Thirdly, we will focus on the E2E aspects including architecture as well as the orchestration and the stitching of the individual domains including data models, data streaming, and verification.