论文部分内容阅读
按照概率密度函数形状,给出了一种比较通用的非高斯噪声模型——多模噪声。多模噪声总体上属于非高斯噪声,但兼容了高斯噪声。改进了高阶统计量的双谱算法,给出一种基于双谱的多模噪声中信号的检测方法,并在此基础上结合无惯性非线性变换器和双谱技术,改进了传统的自适应幅频干扰抑制器,可以精确估计或检测信号。仿真表明该方法可以抑制高斯噪声,同时在强噪声和复杂背景下可以以较高的检测概率检测出信号,优于传统的似然比检测。
According to the shape of probability density function, a more general non-Gaussian noise model is given - multimode noise. Multimode noise is generally non-Gaussian noise, but is compatible with Gaussian noise. The bispectrum algorithm of high-order statistics is improved, and a method of signal detection based on bispectrum multi-mode noise is given. Based on this, a novel method of combining the non-inertial nonlinear transformer with bispectrum is proposed to improve the traditional self- To adapt to the amplitude and frequency interference suppressor, you can accurately estimate or detect the signal. The simulation results show that this method can suppress Gaussian noise, and can detect the signal with higher detection probability under strong noise and complex background, which is better than the traditional likelihood ratio detection.