Periodic Reporting for period 1 - 6G-EWOC (AI-Enhanced fibre-Wireless Optical 6G network in support of connected mobility)
Reporting period: 2024-01-01 to 2025-06-30
The sixth generation of wireless communication (6G) promises transformative opportunities across many sectors, particularly in intelligent transportation. The 6G-EWOC project contributes to this vision by focusing on autonomous driving and connected mobility. Its architecture integrates optical wireless communication (OWC), advanced sensing, and artificial intelligence (AI) to enable vehicles and infrastructure to share information instantly, enhancing both safety and efficiency on the road.
6G-EWOC combines several key innovations:
Optical Wireless Communication and Visible Light Communication (VLC): Delivering high-speed, interference-free connectivity up to 10 Gb/s for vehicle-to-vehicle and vehicle-to-infrastructure links.
Connected LiDAR and RaDAR: Supporting joint sensing and communication for real-time situational awareness.
High-Capacity Fronthaul: Photonic Integrated Circuits (PICs) and ASICs enable cost-effective 50–100 Gb/s coherent fiber connectivity to edge datacenters.
SDN-Enabled Photonic Switching: Providing ultra-low-loss optical switching with microsecond reconfiguration.
AI-Orchestrated Control: Using SDN to dynamically allocate resources, optimize energy efficiency, and enable responsive network management.
Crowdsourced Sensor Fusion: Achieving accurate environmental mapping (SLAM) and cooperative perception across diverse sensor inputs.
These technologies address critical use cases:
Safety: Assisting vehicles in navigating complex environments and detecting hidden objects.
Efficiency: Reducing congestion and emissions through coordinated traffic management.
Emergency Response: Enabling priority routing for emergency vehicles.
Protection of Vulnerable Road Users (VRUs): Enhancing detection and alerts for pedestrians, cyclists, and e-scooter users.
Demonstrations have validated core capabilities, including 10 Gb/s Fi-Wi-Fi optical bridges, long-range VLC-based vehicle-to-vehicle communication, and AI-driven LiDAR/RaDAR sensor fusion for dynamic object detection.
A forthcoming 6G testbed will interconnect UPC (Barcelona), CTTC (Castelldefels), and OTE Labs (Athens) via optical links, supporting the development of smart mobility services and testing on an instrumented, connected vehicle at UPC (Terrassa).
6G-EWOC combines several key innovations:
Optical Wireless Communication and Visible Light Communication (VLC): Delivering high-speed, interference-free connectivity up to 10 Gb/s for vehicle-to-vehicle and vehicle-to-infrastructure links.
Connected LiDAR and RaDAR: Supporting joint sensing and communication for real-time situational awareness.
High-Capacity Fronthaul: Photonic Integrated Circuits (PICs) and ASICs enable cost-effective 50–100 Gb/s coherent fiber connectivity to edge datacenters.
SDN-Enabled Photonic Switching: Providing ultra-low-loss optical switching with microsecond reconfiguration.
AI-Orchestrated Control: Using SDN to dynamically allocate resources, optimize energy efficiency, and enable responsive network management.
Crowdsourced Sensor Fusion: Achieving accurate environmental mapping (SLAM) and cooperative perception across diverse sensor inputs.
These technologies address critical use cases:
Safety: Assisting vehicles in navigating complex environments and detecting hidden objects.
Efficiency: Reducing congestion and emissions through coordinated traffic management.
Emergency Response: Enabling priority routing for emergency vehicles.
Protection of Vulnerable Road Users (VRUs): Enhancing detection and alerts for pedestrians, cyclists, and e-scooter users.
Demonstrations have validated core capabilities, including 10 Gb/s Fi-Wi-Fi optical bridges, long-range VLC-based vehicle-to-vehicle communication, and AI-driven LiDAR/RaDAR sensor fusion for dynamic object detection.
A forthcoming 6G testbed will interconnect UPC (Barcelona), CTTC (Castelldefels), and OTE Labs (Athens) via optical links, supporting the development of smart mobility services and testing on an instrumented, connected vehicle at UPC (Terrassa).
Objective 1: Optical Wireless Communication for V2V and V2I
Progress has been made in enabling high-rate optical wireless links between vehicles and infrastructure. Vehicle-to-vehicle communication beyond 100 m has been demonstrated using high-flux LEDs, supporting real-time LiDAR point-cloud transmission. Vehicle-to-infrastructure links have reached 1 Gb/s over 50 m using beamformer technology, with potential for 10 Gb/s at longer distances as chip-scale beamformers mature. A fiber–wireless–fiber bridge has also been validated over 63 m, showing 10 Gb/s performance with strong resilience to misalignment and turbulence.
Objective 2: Connected LiDAR and RaDAR for Joint Sensing and Communication
An enhanced MEMS-based LiDAR prototype has been built and tested, operating both in sensing mode (camera-like field of view with dense point clouds) and in communication mode with free-space optical transmission up to 200 m. In parallel, vehicular radar operating at 77–79 GHz has been tested, achieving detection ranges up to 150 m and demonstrating combined sensing and communication. Current throughput is below the 1 Gb/s target but ongoing optimization and laboratory evaluation aim to close this gap.
Objective 3: Tunable Transmitters and Receivers for 50–100 Gb/s Fronthaul
Work is advancing on quasi-coherent ASICs and photonic integrated circuits. First-generation 50 Gb/s ASICs are under test, with second-generation 50 Gb/s and first-generation 100 Gb/s ASICs already designed. Integrated quasi-coherent receiver photonic chips with 40 GHz and 100 GHz bandwidth photoreceptors have been developed, with wafer bonding, backend fabrication, and packaging underway. Integration of a tunable on-chip laser remains in progress.
Objective 4: SDN-Enabled Photonic Switching
A new generation of optical switches is being developed to extend software-defined networking down to Layer 1. A 16×16 switch matrix has been fabricated on a low-loss platform and is now entering testing with dedicated driver electronics. On the software side, time-slotted operation is being validated, ensuring reliable, out-of-order-free transmission at sub-microsecond reconfiguration speeds.
Objective 5: AI-Assisted Network Control
An AI-powered SDN control system is under development to optimize the programmability and energy efficiency of 6G-EWOC networks. Device-level power models and energy-aware routing have already demonstrated tangible reductions in power use. End-to-end service provisioning has been shown in testbeds in under 10 seconds, and efforts are ongoing to adapt this for quasi-coherent transceivers and optical switches. A heuristic routing algorithm is in place, while deep reinforcement learning is being pursued for further energy efficiency gains.
Objective 6: AI Applications for Autonomous Vehicles
The project advances Cooperative Perception, enabling vehicles and infrastructure sensors to collaboratively build accurate environmental reconstructions. A vehicle-mounted sensor suite is operational, while roadside multimodal sensors are being deployed. Low-parallax data fusion has been achieved, and work is progressing on spatial registration and time synchronization. The Cooperative Perception algorithm already outperforms state of the art, reaching over 90% accuracy in synthetic datasets and strong gains in real-world data.
Progress has been made in enabling high-rate optical wireless links between vehicles and infrastructure. Vehicle-to-vehicle communication beyond 100 m has been demonstrated using high-flux LEDs, supporting real-time LiDAR point-cloud transmission. Vehicle-to-infrastructure links have reached 1 Gb/s over 50 m using beamformer technology, with potential for 10 Gb/s at longer distances as chip-scale beamformers mature. A fiber–wireless–fiber bridge has also been validated over 63 m, showing 10 Gb/s performance with strong resilience to misalignment and turbulence.
Objective 2: Connected LiDAR and RaDAR for Joint Sensing and Communication
An enhanced MEMS-based LiDAR prototype has been built and tested, operating both in sensing mode (camera-like field of view with dense point clouds) and in communication mode with free-space optical transmission up to 200 m. In parallel, vehicular radar operating at 77–79 GHz has been tested, achieving detection ranges up to 150 m and demonstrating combined sensing and communication. Current throughput is below the 1 Gb/s target but ongoing optimization and laboratory evaluation aim to close this gap.
Objective 3: Tunable Transmitters and Receivers for 50–100 Gb/s Fronthaul
Work is advancing on quasi-coherent ASICs and photonic integrated circuits. First-generation 50 Gb/s ASICs are under test, with second-generation 50 Gb/s and first-generation 100 Gb/s ASICs already designed. Integrated quasi-coherent receiver photonic chips with 40 GHz and 100 GHz bandwidth photoreceptors have been developed, with wafer bonding, backend fabrication, and packaging underway. Integration of a tunable on-chip laser remains in progress.
Objective 4: SDN-Enabled Photonic Switching
A new generation of optical switches is being developed to extend software-defined networking down to Layer 1. A 16×16 switch matrix has been fabricated on a low-loss platform and is now entering testing with dedicated driver electronics. On the software side, time-slotted operation is being validated, ensuring reliable, out-of-order-free transmission at sub-microsecond reconfiguration speeds.
Objective 5: AI-Assisted Network Control
An AI-powered SDN control system is under development to optimize the programmability and energy efficiency of 6G-EWOC networks. Device-level power models and energy-aware routing have already demonstrated tangible reductions in power use. End-to-end service provisioning has been shown in testbeds in under 10 seconds, and efforts are ongoing to adapt this for quasi-coherent transceivers and optical switches. A heuristic routing algorithm is in place, while deep reinforcement learning is being pursued for further energy efficiency gains.
Objective 6: AI Applications for Autonomous Vehicles
The project advances Cooperative Perception, enabling vehicles and infrastructure sensors to collaboratively build accurate environmental reconstructions. A vehicle-mounted sensor suite is operational, while roadside multimodal sensors are being deployed. Low-parallax data fusion has been achieved, and work is progressing on spatial registration and time synchronization. The Cooperative Perception algorithm already outperforms state of the art, reaching over 90% accuracy in synthetic datasets and strong gains in real-world data.
6G-EWOC has generated numerous results that advance beyond the current state of the art, as demonstrated by over 25 publications in scientific journals and international conferences. Due to the limited space available in this section, it is challenging to highlight all the most relevant contributions. Therefore, we encourage readers to explore the full collection of published state-of-the-art results through the following resources:
- Publications Section of the 6G-EWOC Website:
https://6g-ewoc.eu/publications-scientific-papers/(opens in new window)
- ZENODO 6G-EWOC Community:
https://zenodo.org/communities/6g-ewoc/records?q=&l=list&p=1&s=10&sort=newest(opens in new window)
- Publications Section of the 6G-EWOC Website:
https://6g-ewoc.eu/publications-scientific-papers/(opens in new window)
- ZENODO 6G-EWOC Community:
https://zenodo.org/communities/6g-ewoc/records?q=&l=list&p=1&s=10&sort=newest(opens in new window)