Additive random sampling for radar signal processing

This paper gives an overview of challenges and opportunities in digital non-equidistant signal processing. It is pointed out that contrary to common thought non-equidistant signal processing might be of crucial interest to application, bringing important advantages for harmonics detection without having to consider Nyquist restrictions. Main differences between equidistant and non-equidistant sampling with regard to Discrete Fourier Transformation (DFT) of the samples are investigated. In this context, also important sampling schemes are presented. Based on this, we propose four different sampling schemes and investigate the aliasing in frequency domain. The influence of the different sampling schemes on spectral repetitions is demonstrated. Finally, detection rates (including the detection of the correct frequency) for a false alarm rate of 10
⁻⁴
are presented.