Resolution enhancement with model-based frequency estimation algorithms in radar signal processing

In this work we present results on the improvement of resolution capability and accuracy for radar signal evaluation by means of model-based frequency estimation algorithms. In frequency modulated continuous wave radar sensors, which are widely used in industrial contactless distance measurement applications, the usage of the Fourier Transformation for signal evaluation is very common. Nevertheless, using the Fast Fourier Transformation, the resolution capability for closely spaced targets is limited and directly related to the employed signal bandwidth. Model-based frequency estimation algorithms, developed during the last decades, can essentially improve target resolution and distance accuracy. These improvements are shown on simulation data as well as on measured data. Model order estimation, still a challenge when applying model-based evaluation techniques, is tackled by an adaptive approach.