| Direct sequence spread spectrum communication systems has been very widely used in the military and civilian areas, with the implementation of 3G technology in the mobile communication area these two years, DSSS technology is becoming more and more important. Of course, the research issues of interference suppression in DSSS communication systems was highly regarded by scholars at home and abroad.The main interference of DSSS systems are narrowband interference, wideband interference and pulse interference, and the research of narrowband interference suppression has been relatively mature, while the problem of broadband interference suppression has not been solved. Linear frequency-modulated interference is a kind of broadband non-stationary interference, which has been used in the areas of radar, sonar, biomedical and communications, in addition, it is a typical signal that used to measure whether a time-frequency analysis method is effective or not. The problem of LFM interference suppression in DSSS systems is the core of this paper.So far, the main technologies that is used to suppressed LFM interference in DSSS systems at home and abroad are adaptive prediction filtering technology and transform domain interference suppression technology.However, the adaptive prediction filtering technology exists deficiencies as follows:convergence time is slow and lack of matched filter performance,while transform domain interference suppression technology with superior performance. The main transform domain techniques are:Short Time Fourier Transform (STFT), Discrete Chirp-Fourier Transform (DCFT), Wavelet Transform, Gabor Transform, Radon-Wigner Transform, Fractional Fourier Transform (FRFT). STFT is a linear transform, so the calculation is greatly reduced, and it does not exist cross-term interference problem when dealing with polynomial signal, which is has brought great convenience for signal processing. However, the window function of STFT can only change its position, and its size can not be changed, which lead to the contradiction between time and frequency resolution,also, STFT technique is not suitable for dealing non-stationary signal;DCFT technology is also a linear transform,it has no cross terms for interference signal with multi-component LFM signal even if exist the interference plus noise, so this method has a strong noise immunity.In addition, this method has a clear physical meaning and simple calculation, it is effective to estimate and detect LFM signals specially in low SNR and can achieve better bit error rate performance. Finally, you can use fast Fourier transform (FFT) to achieve DCFT, greatly improving the processing speed. MDCFT has a better matching performance compared to DCFT for dealing LFM interference signal.However, MDCFT has a weaker characteristics in terms of the energy than DCFT, but if in non-ideal circumstances, that is to say, if there is not an accurate sampling clock case in the receiver, so the initial frequency and chirp rate of LFM interference is no longer Integers, then it means there will be leakage of energy, and this algorithm can not accurately estimate the parameters; FRFT is especially suitable for processing LFM interference signals. FRFT is a linear transform, that is, there is no cross-interference between multiple components. However, there are some limitations and drawbacks about the algorithms for solving DFRFT, so there must be a find a big limitation about the application of FRFT before we can find a better algorithm for solving DFRFT. In addition, on many occasions, we need to use sweep surface algorithms or repeated operations detection algorithm for FRFT, which are very difficult to calculate and we has yet to find a more suitable faster algorithm.For the problem of LFM interference suppression in DSSS systems, this paper proposed a time-frequency analysis method which based on LPTFT, and then make improvements to this method, the improved algorithm has a small computation, high precision. |
PDF Full Text Request