5GMED Project : innovation, connectivity, challenge. • IRT Saint Exupéry • Technological Research Institute

After a year and a half, discussions between the 21 partners have led to the definition of use cases with their associated services, targeted performances and the necessary infrastructure. The 5GMED project team continues to work on the development of the infrastructure and components that will allow to test the defined services on the corridor. The results of 5GMED are intended to enable the widespread deployment of high-speed, low-latency services for trains and cars on all cross-border corridors in Europe.

THE PROJECT

The objectives of the project

The 5GMED project was launched in September 2020 for a duration of 3 years and a budget of €15 million.

The objective of the project is to test on a 40 km cross-border corridor (freeway + railroad) – between France and Spain – innovative services based on the 5G telecom network and AI, with a particular focus on service continuity when roaming between the two countries.

5GMED is part of a series of European projects that aim to generalize the deployment of high-speed, low-latency services for trains and cars on all cross-border corridors in Europe.

The infrastructures

The services experienced within 5GMED project are based on the deployment of the following infrastructures :

A 5G telecommunication infrastructure deployed along the corridor in France, and as a densification of Vodafone network in Spain. Vodafone is one of the project’s partners. Several of its 4G sites will be reused to deploy the 5G access network in the Country. Most of these new sites deployed will be energy self-sufficient as far as possible. Backhauling* of some sites will be implemented via a satellite link.
Deployment of other radio access networks, non 3GPP: IEEE 802.11p, IEEE 802.11 ad, satellite.
A computing infrastructure in the cloud. Some applications related to the services tested by 5GMED will be deployed in the cloud.
A computing infrastructure on the “EDGE”, i.e. along the corridor, allowing to obtain the low latency required by some services, by bringing some of the network functions or applications closer to the user. This computing capacity will enable the deployment of AI* modules used by certain services.

*Backhauling: the action of transmitting and receiving data between the radio access network and the core network ; *AI: Artificial Intelligence

FOR THE RAILWAYS

Highly innovative services will enable the testing of high-speed, low-latency transmissions from the train to the ground and vice versa. These services will be seamless along the entire corridor and especially at border crossings (roaming between the two mobile network operators):

Massive transmission of data from sensors located in the train (mIoT) to the train operator’s IS*
Obstacles detection using LiDAR technology (Light Detection And Ranging). Lidar data are processed either in the train or on the EDGE by an AI module;
Seamless and high speed WiFi;
On-board security: transmission of video streams from CCTV cameras located in the train cars to the train operator in case of manual alarm or alarm detected by an AI module (permanent analysis of video streams by an AI located on the ground);
Mobile backhauling of a neutral host 5G cell located in the train. This cell is deployed in the train by a third party infrastructure provider and will allow passengers, entering the train, to roam on this cell. This will ensure 5G coverage in the train, independently of the mobile network operators;
A service called EMT (Enjoy Media Together) allowing the transmission of HD video streams to the train or to the cars in synchronization. This service is optimized by an AI module.

*IS: Information System

FOR THE AUTOMOTIVE

As for the railways, very innovative services will allow to test high speed and low latency transmissions from the car to the ground and vice versa. These services will be seamless on the whole corridor and especially at the border crossing (roaming between the two mobile network operators).

Experimentation of a remote vehicle teleoperation service: an operator located on the ground takes control of a vehicle and drives it from point A to point B on the highway. For example, a vehicle on the emergency lane can be remotely operated to the next highway rest area. This service is optimized by an AI module;
The EMT service will also be deployed for automotive applications;
Relay of emergency messages sent by a vehicle to other vehicles in the local area;
Automatic incident detection and regulation of local traffic;
Automatic regulation of traffic on whole portions of the highway, based on a small subset of connected and autonomous vehicles among the classic non-connected vehicles.

THE ROLE OF THE IRT SAINT EXUPERY

The IRT Saint Exupéry will use its expertise in 5G and satellite networks as well as in AI to develop a satellite slice, but also several AI modules.

The IRT Saint Exupéry has been responsible for the initial definition of services and testing for each use case. It will also use its expertise in 5G and satellite networks as well as in AI to develop:

The obstacle detection AI module in collaboration with the NPM3D laboratory of Paris’ Ecole des Mines;
The AI module for orchestrator optimization. In interface with the orchestrator, it allows an optimal placement of the virtual functions at the EDGE and, thus, ensures the continuity of the EMT service performances along the corridor (throughput, latency, etc.);
The QoS prediction AI module for the remote vehicle operation service on the highway;
The development and testing of a satellite slice at the IRT Saint Exupéry before its implementation in the train. This slice is integrated into the overall telecom architecture deployed in the train.