Starting date : May 2018 > Apr. 2021
Lifetime: 36 months
Program in support : ECSEL
Status project : complete
CEA-Leti's contact :
Sylvie Mayrargue
Bernard Strée
Project Coordinator: NXP Semiconductors Netherlands BV (NL)
Partners: - AT: AIT, AVL, Cisc Semiconductors, Secinto, Thalès Austria, Virtual Vehicle
- BE: IMEC, Transport & Mobility Leuven,
- CZ: Budapest University of Technology, Institute of Microelectronic Applications
- DE: AVL, Fraunhofer, Giesecke+Devrient Mobile Security, Merantix, NXP Semiconductors, Roche, Senetics Healthcare Group, Technical University of Kaiserslautern, ZF Friedrichshafen
- ES: Advanced Automotive Antennas FICO, FICOSA ADAS, INDRA Systems, CSIC,TST,
- FI: Haltian, Nokia, Solita, University of Oulu
- FR: Canon Research Centre France, CEA-Leti, Gemalto, IDEMIA, IFSSTAR, In2car, Internet of Trust, ISSM Invia, PSA, Prove and Run, YoGoKo
- GB: NXP
- HU: Commsignia, University of Budapest
- IT: Evidence, Ideas & Motions, Magneti Marelli, University of Modena & Reggio Emilia
- NL: Catholic University of Nijmegen, Fastree3D, IMEC-NL, Municipality of Helmond, Philips, TNO, Ubiqu Access, Technical University of Eindhoven
- PL: Technical University of Gdansk
- PT: Beyond Vision, Infrastructures of Portugal, Institute of Telecommunications, IP Telecom, PDM&FC
- RO: University Polytechnical of Bucarest
- SE: RISE, China-Euro Vehicle Technology, Technology Nexus Secured Business Solutions
- TN: ENIT, SUPCOM
Target market: n/a
Publications:
«Localization and Communication Resource Budgeting for Multi-user mmWave MIMO», R. Koirala, B. Denis, B. Uguen, D. Dardari, H. Wymeersch, IEEE Workshop on Positioning, Navigation and Communications 2019 (IEEE WPNC’19), Bremen, Oct. 2019.
Investment: € 51.5 m.
EC Contribution: € 51.5 m.
| Stakes
Radars and lidars are two of the main sensors enabling car autonomy in a globally safer ecosystem. Their raw data need to be processed to obtain Occupancy Grids (OGs), i.e. a map of obstacles (fixed or mobile) surrounding the vehicle. However, on-board calculations must not be over-complex in order to save battery energy. CEA-Leti has developed an approach for computing OGs from RADAR measurements. The resulting OG will serve as an input for a lightweight data fusion algorithm that will enable LIDAR and RADAR data to be combined to create a map of the vehicle surroundings.
Automotive radars are expected to operate in the millimeter wave (mmWave) frequency domain so CEA-Leti has also investigated various aspects of their integration within a communication-radar-localization multi-service framework, which relies on V2X wireless links to other connected vehicles and/or fixed road infrastructure elements. Based on characterization of theoretical performance bounds in multi-user and multi-carrier contexts, CEA-Leti initially studied how optimized beamforming could enhance estimation of intermediate location dependent radio variables (e.g. delay and angles of departure/arrival of transmitted mmWave signals) and thus moving vehicle position/orientation. Non-trivial operating trade-offs were then highlighted among these intricate services (typically between spatial resolution, rate coverage and latency). As a result, several resource allocation schemes have been proposed (over time/frequency/users) along with location-based beam optimization/alignment strategies. Finally, by exploiting multipath departure/arrival angles in sparse mmWave channels, CEA-Leti has considered applying simultaneous localization and mapping (SLAM) algorithms to calculate the positions of moving vehicles and passive scattering obstacles in the vicinity, thus paving the way towards opportunistic radar functions over V2X communication links.
SECREDAS aims to develop and validate multi-domain architecting methodologies, reference architectures and components for autonomous systems by combining high security and privacy protection, while preserving functional safety and operating performance.
IMPACT
The SECREDAS project contributes to ensuring faster deployment of autonomous/connected cars by developing appropriate enabling technologies.
Radars and lidars are key to autonomous car development. However, onboard processing of their measurement data drains the car battery. It is therefore of utmost importance to reduce the energy consumption of the relevant data fusion algorithms. Future radars will operate in the millimeter spectrum (79GHz) so vehicle communication (between cars and/or car/infrastructure) at these frequencies can also be used to ensure accurate, opportunistic localization of surrounding vehicles and/or obstacles, thereby providing information additional to that from conventional radars and lidars.
|
|