Time-frequency Zadeh filtering in estimation of radar signals embedded in non-gaussian disturbances

摘要

In this paper, the Zadeh method for the time-varying filtering of the non-stationary radar signal, embedded in non-gaussian disturbances, has been considered analytically and numerically. The non-stationary radar signal has been received from a real radar receiver and only a real realisation has been accessible for experiments. The additive noise signal with Weibull distribution, as en example of non-gaussian disturbances, has been taken into account. The linear, time-varying (LTV) Zadeh filter, for suppressing the noise signal and passing the radar signal, has been developed based on the concept of the Weyl symbol. The Weyl symbol, or equivalently the indicator function, has been designed from the Short Time Fourier Transform (STFT) as the time-frequency (TF) representation of a noisy signal. Comprehensive numerical experiments illustrate the results of the filtering process.