Cloud Native 5G: an Efficient Orchestration of Cloud Native 5G System
Abderaouf Khichane, Ilhem Fajjari, Nadjib Aitsaadi, Mourad Gueroui
Abstract
Cloud native paradigm gained momentum during the last few years fostering its adoption in 5G architecture. In this context, a new generation of network functions called Cloud native Network Functions (CNFs) has seen the light of day. However, despite their numerous advantages in terms of lightness and portability, CNFs raise new issues not solved yet. Specifically, these fine-grained services require an efficient orchestration system able to automate their lifecycle management while considering the stringent QoS’s requirements. In this paper, we propose a novel 5G CNFs orchestration framework addressing both IT and Network resource provisioning. Both qualitative and quantitative studies are conducted to validate our solution using the auto-scaling use-case. Our extensive experimentations show that our proposal achieves good performances in terms of: i) deployment time, ii) upgrade time, iii) packet loss rate, and iv) resource allocation balancing.