Nonparametric Spectral Estimation of Phase Noise in Modulated Signals Based on Complementary Autocorrelation
Nonparametric Spectral Estimation of Phase Noise in Modulated Signals Based on Complementary Autocorrelation
复制标题
基于互补自相关的调制信号相位噪声非参数谱估计
DOI：
10.1109/tsp.2014.2336606
复制
发表时间：
2014
影响因子：
5.4
通讯作者：
K. Harada
中科院分区：
文献类型：
--
作者：
K. Harada
提出了一种基于互补自相关（CAC）的相位噪声非参数谱估计方法。利用相位噪声的CAC不为零的事实，该技术提供了一个明显的优点，即相位噪声的功率谱被估计，而其他类型的噪声的功率谱被抑制，只要它们是适当的（二阶循环性）。当直接从数字调制信号估计相位噪声频谱时，该特性最有用，因为这些信号通常伴随着各种不期望的噪声，这些噪声容易掩盖感兴趣的相位噪声。针对相位噪声谱的实际计算，提出了一种基于互补谱的频域估计方法。对于该方法，在假设噪声适当的情况下，推导出了噪声抑制因子的计算公式。此外，可抑制性分析进行，这表明，许多与调制信号相关的噪声确实是适当的，因此可以被抑制。数值仿真验证了该方法的有效性，并验证了噪声抑制因子。该技术首次实现了在提取相位噪声谱的同时抑制多种噪声。
A novel nonparametric spectral estimation technique for phase noise based on the complementary autocorrelation (CAC) is presented. Utilizing the fact that the CAC of phase noise is not zero, this technique offers a distinct advantage that the power spectrum of phase noise is estimated whereas that of other kinds of noise are suppressed as long as they are proper (second-order circularity). This property is most useful when estimating the phase noise spectrum directly from digitally-modulated signals because those signals are typically accompanied by various kinds of undesired noise that easily mask the phase noise of interest. A frequency-domain estimation method based on complementary spectrum is presented for practical computation of phase noise spectrum. For the method, a formula for the noise suppression factor is derived assuming the properness in noise. Furthermore, suppressibility analyses are conducted, which reveal that many of the noises associated with modulated signals are indeed proper, and can thus be suppressed. Numerical simulations demonstrate the method, and verify the noise suppression factor. It is believed that this technique is the first of its kind being able to extract phase noise spectrum while suppressing many other kinds of noise.