Low-Latency Service Schedule Orchestration in NFV-based Networks
Hyame Assem Alameddine, Chadi M. Assi, Mosaddek Hossain Kamal Tushar, Jia Yuan Yu · 2019
The Fifth Generation (5G) era is bringing tremendous new network capabilities enabling diverse services belonging to different business verticals (i.e., manufacturing, automotive, etc.) and provided with top-notch Quality of Service (QoS) (i.e., ultra-low latency, ultra-reliability, etc.). Empowered by soft-warization technologies such as Network Function Virtualization (NFV), 5G networks are envisioned to be agile, sustainable and self-organized. NFV promotes the automated provisioning of Network Services (NSs) through processing their traffic by a chain of Virtual Network Functions (VNFs). As VNFs are shared between multiple NSs, a clear approach to map and schedule the carried traffic of these services is required. Hence, in this paper, we solve the Latency-Aware Service Schedule Orchestration problem (LASSO) that jointly addresses the mapping and scheduling of services to VNFs. We formulate the problem as a Mixed Integer Linear Program (MILP) and we present ENCHAIN, a novel game-theoretic approach exploiting a scalable solution for the LASSO problem while providing each NS the freedom to decide on its own mapping and scheduling solution.