JultikaUniversity of Oulu repository

Abstract

5G Network has many important properties, including increased bandwidth, increased data throughput, high reliability, high network density, and low latency. This thesis concentrate on the low latency attribute of the 5G Standalone (SA) mode and 5G Non-Standalone (NSA) mode.

One of the most critical considerations in 5G is to have low latency network for various delay-sensitive applications, such as remote diagnostics and surgery in healthcare, self-driven cars, industrial factory automation, and live audio productions in the music industry. Therefore, 5G employs various retransmission algorithms and techniques to meet the low latency standards, a new frame structure with multiple subcarrier spacing (SCS) and time slots, and a new cloud-native core.

For the low latency measurements, a test setup is built. A video is sent from the 5G User Equipment (UE) to the multimedia server deployed in the University of Oulu 5G test Network (5GTN) edge server. The University of Oulu 5GTN is operating both in NSA and SA modes. Delay is measured both for the downlink and the uplink direction with Qosium tool. When calculating millisecond-level transmission delays, clock synchronization is essential. Therefore, Precision Time Protocol daemon (PTPd) service is initiated on both the sending and receiving machines.

The tests comply with the specifications developed at the University of Oulu 5GTN for both the SA and the NSA mode. When the delay measurement findings were compared between the two deployment modes, it was observed that the comparison was not appropriate. The primary reason for this is that in the 5GTN, the NSA and the SA have entirely different data routing paths and configurations. Additionally, the author did not have sufficient resources to make the required architectural changes.