System Design And Signal Processing Of Doppler Through-the-wall Radar

The study of the human body movement micro-doppler signal from a throughthe-wall radar is a hot spot in through-the-wall radar signal processing. Becausedifferent parts of the body make different radar wave modulations, the human bodymovement micro-doppler signal from a through the-wall radar is a non-stationary signal.By processing the non-stationary signal with time-frequency analysis method, theinformation of the signal frequency changing over the time can be obtained. The state ofhuman motion behind the wall will be better judged with this kind of information.As the traditional time-frequency analysis method, short-time Fourier transform andwavelet transform are uesd in many applications. Both of the methods can demonstratethe frequency distribution in time of the non-stationary signal. The essence of short-timeFourier transform is a windowed Fourier transform, if there is a clear physical meaningof this transform, the analyzed signal should be local stable, but the measured datausually do not have such a nature. Wavelet transform is similar to short-time Fouriertransform, its algorithm is also based on the convolution of the basis function with thesignal. Both methods are troubled by the uncertainty principle and the selecting of basisfunction, and do not have good time-frequency resolution.As an adaptive time-frequency analysis method, Hilbert-Huang transform can notbe affected by uncertainty principle and is playing an increasingly important role innon-stationary signal processing. This paper discusses the theories of short-time Fouriertransform, wavelet transform and Hilbert-Huang transform, and designs a simplecontinuous wave signal-frequency through-the-wall radar transmitting and receivingsystem. The three methods mentioned are used in the analysis of through-the-wall radarhuman motion signal. In order to solve the problem of false component and modalaliasing existed in through-the-wall radar human motion signal when usingHilbert-Huang transform, the algorithm has been improved. The through-the-wall radarhuman motion signal from the through-the-wall radar system is uesd to test theimproved algorithm and good result is got.