Real-time Simulation of Multiple Connected Autonomous Vehicles
Connecting vehicles (car-to-car C2C, car-to-infrastructure C2I, and applying mobile IoT) will significantly contribute to safe and secure autonomous driving in complex road-traffic scenarios. As an example, the automation of inner-city traffic will immensely benefit by being able to use extended range and complementary situational awareness in combination with smart city (IoT) concepts.
The goal of project REALISM is to provide a software framework for running highly parallelized real-time simulation on commercial high-performance PC’s with 20 and more CPU cores. This will enable academia and industry to analyse, develop and validate complex communication scenarios of mobile systems in real time, allowing mixed virtual (simulated) and real (physical devices) setups in XiL (“any-in-Loop”) setups. REALISM will therefore provide novel physical-layer system-level simulation models for C2X communication of multiple vehicles, optimized for parallel execution on multi-core platforms, and a demonstrator to validate the performance of the multi-core framework running the described simulation in a closed-loop setup, including transmitting simulation results via radio signals in real time to a test vehicle within an automotive vehicle test cell.