Flexible MEC service consumption through edge host zoning in 5G networks

TL;DR

This paper proposes a latency-aware zoning approach for MEC services in 5G networks, improving end-to-end performance by considering proximity to MEC hosts, especially for control plane traffic in virtualized environments.

Contribution

It introduces a novel proximity zone concept based on statistical processing time models to enable flexible MEC service consumption across different localities.

Findings

Latency-aware proximity zones improve E2E performance.

Proximity zones enable flexible MEC service consumption.

Performance benefits are demonstrated in exemplary scenarios.

Abstract

Multi-access Edge Computing (MEC) is commonly recognized as a key supporting technology for the emerging 5G systems. When deployed in fully virtualized networks, i.e., following the Network Function Virtualization (NFV) paradigm, it will enable a multitude of new applications and use cases. However, the growing number of devices, combined with the vastly increasing traffic demand, call for low End-to-End (E2E) latency packet transfer and processing in an NFV environment, both in user and control plane. In this paper, focusing on control plane packet traffic, we investigate the general case of a MEC application consuming a MEC service running on a different MEC host. To enable flexible MEC platform service consumption at different localities, based on a state-of-the-art statistical model of the total processing time, we define latency-aware proximity zones around MEC servers hosting MEC application instances. Exemplary scenarios exhibit the E2E performance benefit of introducing the awareness of proximity zones around MEC hosts and service producing MEC application instances. This performance-aware service consumption will be beneficial in view of the future evolution towards distributed computing systems.