Moreno Leon, CarlosCarlosMoreno LeonGonzález-Huici, María A.María A.González-HuiciDallmann, ThomasThomasDallmann2022-03-142022-03-142019https://publica.fraunhofer.de/handle/publica/41030610.1109/ICMIM.2019.8726648One important aspect in the design of a generic radar target simulator is the level of complexity incorporated in the generation of scenarios. The trade-off between the expected fidelity in the generation of scenarios and the computational constraints in the target simulation system raises alternatives to model-based approaches. In this paper we present a data-driven method for the generation of road scenarios in the context of automotive radar target simulation. The method characterizes the scenarios relying on radar recordings and prior information on the testing set-up. The recorded data is processed in order to play back the scenario with a radar target simulator. This is relevant, for instance, so that certain functionalities of the radar under test can be evaluated and tuned in reproducible conditions. The presented data-driven method is applied to one particular road scenario and validated in simulation experiments.en621conference paper