Deutsche Telekom, Nokia tests show edge computing is only way to make 5G-V2X safe
5G networks will only achieve the required reliability and latency for autonomous driving if sufficient edge compute functions are brought close to the roadside and architected correctly. Vehicle-to-everything (V2X) communications, between vehicles and with road infrastructure and services, will not enable driving systems to meet the necessary safety standards otherwise.
This is the conclusion of a two-year study carried out on the A9 motorway test track in Germany, involving Continental, Deutsche Telekom, the Fraunhofer Institute, MHP and Nokia. The quintet, collaborating on the Car2MEC project, said multi-access edge computing (MEC) is key technology to keep connected vehicles safe.
In the end, the project showed LTE came up short as a technology for fully-autonomous driving, and that only the localised compute functions carried shorter distances on ultra-reliable low-latency communications (URLLC) networks, enabled with the next generation of standalone 5G, will make its safe.
Karsten Roscher, senior scientist at Fraunhofer ESK, commented: “While the edge cloud improved end-to-end latency by more than 20ms, the 4G radio access network is still a limiting factor for more demanding future services. However, we believe the combination of edge cloud and 5G will shape a completely new landscape of connected and distributed applications.”
The testing confirmed time critical information can be delivered from one car to another in less than 30 ms in an LTE network combined with n edge-based MEC setup.The performance of latency critical and data intense applications like HD positioning and maps improved significantly with MEC infrastructure. In general terms, for lower levels of automation and “distributed applications”, LTE does achieve sufficient latency and reliability, said Nokia.
Uwe Puetzschler, head of car-to-everything at the Finnish vendor, commented: “The results mark a big step forward on the way to safe and automated driving. We have demonstrated that the edge cloud on top of 4G and 5G networks allows to deploy distributed applications that meet the requirements of the automotive industry in terms of latency and reliability. This will accelerate a commercial deployment of the technology.”
The Car2MEC project, funded by the Bavarian Ministry for Economic Affairs, tested a number of different use cases on the A9 motorway, including for emergency warnings, ‘end-of-jam’ warnings, and variable speed limit assistance, and HD maps. It saw German operator Deutsche Telekom deploy additional network infrastructure alongside its public LTE network with two locally separated MEC sites in the test area of the A9 motorway.
The testbed shared resources with a live LTE network, and operated for 12 months. The group said it presented a “unique opportunity to assess MEC technology in a real-world environment”, combining public road and public network infrastructure in live usage.
Alexander Lautz, senior vice president of 5G at Deutsche Telekom, said: “The project results validate the performance of edge computing over our 4G networks as a potential enabler for automotive cases that require low latency and ultra-high reliability.”
Ronald Hain, backend team leader for interior systems and technology at Continental, said: “By closing the gap between local real-time applications and cloud services, MEC will allow us to enhance automated driving and enable vehicles to cooperate with each other. In addition, local services effectively utilise the data rate of LTE or 5G networks.”
The Fraunhofer Institute for Embedded Systems and Communication Technologies (ESK) developed “hybrid communication units enabled time-sensitive applications for connected vehicles with proven, standardised protocol and message sets”.
A press statement said: “A highly efficient distributed messaging service deployed on the multi-access edge nodes provided low-latency connectivity between the vehicles and a direct link to localised services. In combination with Fraunhofer ESK’s adaptive networking algorithms, which select the most suitable communication path at runtime, the complementary strengths of different technologies can be exploited to provide optimised quality-of-service for each application and context.”
MHP considered the business perspective, and described the opportunities MEC offers to the automotive industry are “ambiguous, yet promising”. It said cooperation, partnerships, standardisation are crucial for its development.
Olaf Kleindienst, partner at MHP, said: “Everybody is talking about the connected car and, in some ways, the connected car is already there. But it still has some way to go and many questions remain, particularly around the fully connected car. The fully connected car represents a new environment for the automotive industry and the established key players must reinvent themselves, with new business models.”