Acoustic noise atmospheric radar with nonparametric copula based signal processing

Abstract

In this paper we discuss a new approach to creating the acoustic atmospheric radar (sodar) which is based on noise sounding waveform. Such sounding waveform has extremely high range and speed resolution. These good properties exist only in the case when we have independent samples, and a noise signal forms such independent samples because of its nature. The prototype of the radar is described and investigated theoretically, using computer simulation and in natural conditions. Non-parametrical signal processing algorithm is developed based on the statistical approach and digital signal processing. It has wonderful properties of invariance to the group of the noise and signal transforms and stable level of the false alarm probability. The algorithm can be applied for both acoustic and microwave noise radars. The use of the FFT algorithm increases the speed of calculations and gives us the possibility of constructing digital random signal radars. These new properties allow us to simplify signal detection algorithms and measure a distance, an azimuth and a target velocity simultaneously with high resolution and accuracy.