A Resource Allocation Framework for Network Slicing

Telecommunication networks are converging to a massively distributed cloud infrastructure interconnected with software defined networks. In the envisioned architecture, services will be deployed flexibly and quickly as network slices. Our paper addresses a major bottleneck in this context, namely the challenge of computing the best resource provisioning for network slices in a robust and efficient manner. With tractability in mind, we propose a novel optimization framework which allows fine-grained resource allocation for slices both in terms of network bandwidth and cloud processing. The slices can be further provisioned and auto-scaled optimally based on a large class of utility functions in real-time. Furthermore, by tuning a slice-specific parameter, system designers can trade off traffic-fairness with computing-fairness to provide a mixed fairness strategy. We also propose an iterative algorithm based on the alternating direction method of multipliers (ADMM) that provably converges to the optimal resource allocation and we demonstrate the method's fast convergence in a wide range of quasi-stationary and dynamic settings.

I. Introduction

The infrastructure of telecommunication networks is undergoing a radical transformation towards a massively distributed cloud with network functionalities implemented as virtual network functions (VNFs). When this transition is complete, services will be instantiated as network slices - essentially virtual networks that utilize transport and computing resources in order to provide high-fidelity communications [1]. Unlike their physical counterparts, VNFs can be launched, placed, and scaled flexibly to meet fluctuating workload demands, thus opening new doorways for on-demand resource provisioning. However, quick and accurate resource allocation for multiple network slices is a very challenging task, primarily due to the following reasons: 1)

Each network slice requires specific functionalities, possibly executed in a particular order.

2)

The operation of a network slice involves (and requires the close coordination of) several stakeholders: a) slice owners; b) the cloud provider; and c) the network controller.

3)

Each slice requires two types of resources, bandwidth and processing power. As a result, resource allocation must be balanced between these two components, and notions of fairness and efficiency become much more intricate.