Research On Nonstationary Interference Suppression Based On Time Frequency Analysis

How to suppress non-stationary interference in communication system is a challenging problem in communication counterwork. This paper studies interference excision techniques based on time-frequency analysis tools, such as short time Fourier transform , wavelet packet transform, etc, which is used to suppress non-stationary time-frequency energy concentrated interference patterns. The main contributions are as follows:1 Based on short-time Fourier transform interference suppression, this paper proposes two improved interference suppression methods and applies them to suppress wide-band time-frequency domain concentrated dynamic interference patterns. One improvement rests on the suppression of many spectral lines in order to suppress sweep frequency-spectrum interference, by choosing appropriate short-time Fourier parameters. The other improvement is finding the non-interference spectrum to set the more reasonable interference threshold, according to the same statistical characteristics of spectral lines in each window in one-processing data and combining with amplitude spectrum hard limit , with excellent suppressing result.2 Based on wavelet packet transformation interference suppression, a method based on the wavelet packet coefficient statistical property interference suppression algorithm is proposed. This algorithm dose not need to set energy concentration threshold as the algorithm based on wavelet packet complete decomposition (WPCD), and overcomes the disadvantage of setting threshold by non-interfered leaf nodes in WPCD algorithm. This is helpful to suppress interference occupying the whole spectrums of receiving filter, which makes the BER closes to ideal state, with good suppressing effect.