WONDER: Workload Optimized Network Defined Edge Routing

TL;DR

This paper introduces methods for optimizing traffic routing and workload placement in 5G edge networks, enhancing service performance through intelligent routing and compute distribution across interconnected edge data centers.

Contribution

It proposes novel techniques for traffic routing and workload placement in 5G edge networks using Segment Routing/IPv6 and SDN control mechanisms.

Findings

Improved traffic routing efficiency under mobility conditions.

Enhanced workload distribution across edge data centers.

Better support for high-performance 5G services.

Abstract

The 5G standards enable cellular network capabilities that significantly improve key network characteristics such as latency, capacity, throughput and reliability, compared to the previous generations of wireless networks. It is, however, clear that in order to achieve these improvements in real network implementations, the supporting physical and logical infrastructure needs to be designed appropriately. The key components of this infrastructure are Radio Access Network, Edge Data Centers, Packet/Optical Interconnection Fabric and Edge Computing. This paper concentrates on the Edge Data Centers, Interconnection and Edge Computing capabilities that target ability to deliver high-performing services on the 5G network by means of intelligent network slicing, traffic routing and coordinated compute workload distribution. We propose new methods of ensuring optimal traffic routing and edge compute workload placement under mobility conditions, subject to application requirements and constraints within a set of interconnected Edge Data Centers, utilizing Segment Routing/IPv6 and software defined control mechanisms.