Warnke, MarcelMarcelWarnkeBrüggenwirth, StefanStefanBrüggenwirth2023-07-102023-07-102023https://publica.fraunhofer.de/handle/publica/44539410.1109/TAES.2022.32306592-s2.0-85146239247Waveform adaptation is a key-feature for modern radar systems and essential for cognitive radar. In this work we present a concept for the enhancement of the classification performance by using optimised transmit waveforms and a Gaussian template matching high range resolution profile (HRRP)classifier. A straight forward approach is presented, aiming to improve specific parts of the confusion matrix which will be exploited within a cognitive framework. The optimisation includes different types of uncertainties and is designed during a training process to be accessed by a library. Taking different uncertainties into account, the calculation of the expected performance, the optimisation, the range side lobe constraint and the time-domain realisation is explained. A non-linear frequency modulation (NLFM) waveform is used since it provides a compression gain with range resolution and a constant envelope. Based on an electromagnetic simulation the concept is validated for different ground targets and aspect angle uncertainties. The adaptation is compared to a commonly used linear frequency modulation (LFM). The results of the mean performance improvement reached an enhancement between 8.8% and 20.9%.enClassificationCognitive RadarCognitive radarElectromagneticsGenetic AlgorithmHRRPMeasurementNon-linear frequency modulationOptimizationRadarRadar signal processingSignal to noise ratioUncertaintyWaveform designjournal article