直扩信号盲检测方法研究

Blind Detection of Direct Sequence Spread Spectrum

文中应用谱相关法、倒谱法、高阶累计量法、基于特征值分析等方法对DSSS信号的盲检测进行了分析,总结了各种方法的原理及实现过程,并用Matlab对其检测性能进行了仿真分析,在相同的条件下比较各种方法的检测性能随信噪比和采样点数的变化情况。结果表明,谱相关法对低信噪比的适应能力差,要达到90%的检测概率,所需的信噪比为-6dB。倒谱法、高阶累计量法在采样点数小于4000时,二者的性能和谱相关法相当,随着仿真采样点数的增加,性能提升明显,因此实时性较差。基于特征值分析的方法各方面的性能比较均衡,可以作为一种快速通用检测算法。

In this paper,cyclic spectrum correlation,cepstrum detection,higher-order statistics method,eigenvalue analysis are applied to analyze the detection of DSSS signal.Principles and implementation processes of these methods are summarized.Simulation results show that cyclic spectrum correlation perform poorly at low signal-to-noise ratio（SNR）,and that cepstrum detection and higher-order cumulants cannot do real-time processing.As for the eigenvalue analysis method,it works very well and can be a fast generic detection algorithm.