Automated Driving with 5G Network Slicing and Quality of Service

Deutsche Telekom, BMW, Valeo, Ericsson, and Qualcomm announced the world's first demonstration of an automated driving application supported by 5G Standalone (SA) network slicing with controlled network features for Quality of Service (QoS). Reliable network connectivity is a prerequisite for many emerging automotive use cases. In trials, the partners explored how 5G SA network slicing with different QoS features support end-to-end automotive use case scenarios.

Network slicing is a key feature of 5G SA. Slices can be tailored to applications and use cases with different QoS requirements in terms of performance, security or latency. Using the User Equipment Route Selection Policy (URSP) feature, the device was able to dynamically select and connect to multiple slices simultaneously. In addition, a Network API, which exposes network capabilities, was used to request a defined QoS level from the network. This was requested via a Quality on Demand (QoD) network API. The tests showed the capability of all these features combining to meet the QoS demands of mission-critical applications under various network load situations.

In February 2022 DT, BMW, and Valeo announced the first tests for Quality of Service using the Quality on Demand feature as a network API. The Application Programming Interface (API) was successfully tested for the Automated Valet Parking (AVP) use case and the results were presented at Mobile World Congress (MWC) 2022. At the same time, the CAMARA initiative was announced by GSMA. The CAMARA initiative consists of network operators, technology vendors, cloud providers, OS vendors, and application developers. It aims to standardize network APIs globally. Quality on Demand was the first network API standardized in the CAMARA initiative.

An increasing number of automotive use cases require reliable mobile network connectivity. Use case examples are safety-related features, remote control functions, such as remote door opening/closing, and assisted and automated driving features. The partners set out to test and explore whether and how to offer QoS features in future network deployments. The tests were performed at DT’s testing facilities in Berlin Winterfeldtstraße.

The tests performed demonstrated the high value of QoS features to achievable bandwidth, stability, and latency. The partners showed how 5G SA network slicing supported an end-to-end automotive use case scenario with application-grade connectivity enabled by three key QoS features in the operator network:

Network API - A mission-critical application makes an ad hoc request for an improved network performance via the Quality on Demand (QoD) API feature in an enhanced mobile broadband (eMBB) slice. The trial clearly shows how the application can request and receive network performance via the API to meet its QoS requirements even in congested network situations.

Network Slicing - in the tested scenario mission-critical data was transmitted via a high-quality slice, while non-mission-critical data was transmitted via an eMBB slice. The trial’s measurement results showed that for congested conditions, where many users are sharing the limited resources of the mobile network, the automated driving function was consistently served with the necessary bandwidth resources.

UE Route Selection Policy - in a mission-critical scenario, the URSP feature was used to enable network slice selection on the device side according to network slice availability on the network side. The results demonstrate that application traffic from different applications can be handled by multiple slices in parallel. For example, mission-critical application traffic is transmitted via the high-quality slice, while non-mission-critical application uses the eMBB slice.