论文部分内容阅读
基于傅里叶分析的功率谱密度只能反映 1/f噪声的整体频率特性 ,子波变换模极大值能够反映 1/f噪声的奇异性和非规整性 ,而后者才是 1/f噪声最本质的特征所在 .本文将这一特性用于MOSFET 1/f噪声的相似性分析 .从子波变换模极大值匹配原理出发 ,定义了一个 1/f噪声的相似系数 ,利用它对不同形成机制、不同微观缺陷状态、不同偏置应力作用下的MOSFET 1/f噪声进行了相似性分析 ,发现它可作为鉴别 1/f噪声的物理起源 ,分析 1/f噪声的微观动力学机制 ,筛选有潜在缺陷或损伤的MOS器件的有效手段
The power spectral density based on Fourier analysis can only reflect the overall frequency characteristics of 1 / f noise. The wavelet transform modulus maxima can reflect the singularity and irregularity of 1 / f noise, while the latter is 1 / f noise The most essential feature is that this property is used for the similarity analysis of MOSFET 1 / f noise.According to the principle of wavelet transform modulus maximum matching, a similarity coefficient of 1 / f noise is defined, It is found that it can be used as a physical origin to identify 1 / f noise and analyze the micro-dynamics mechanism of 1 / f noise. An effective means of screening MOS devices that have potential defects or impairments