| 摘要: |
| 依据非线性动力学理论，在音频信号的修正离散余弦变换（MDCT）域，采用自相关法和虚假
近邻法（FNN
）分别计算延迟时间和嵌入维数，重构出音频信号MDCT域信息的相空间，并基于Rosenstein
小数据量法计算最大Lyapunov（李雅普诺夫）指数，依据其正负对音频频域序列的混沌特性
进行了统计分析和验证。实验表明，音频信号MDCT序列的最大Lyapunov指数皆为正，音频信
号具有混沌特性。 |
| 关键词: 音频信号处理 修正离散余弦变换 相空间重构 Lyapunov指数 混沌
特性 |
| DOI： |
|
| 基金项目:国家自然科学基金资助项目(61072089，60872027)；北京市自然
科学基金资助项目(4082006)；北京市属高等学校人才强教计划；北京工业大学国家大学生
创新性实验项目（101000506） |
|
| Chaotic Characteristic Analysis of Audio Signals |
| BAO Feng,LIU Xin,JIA Mao-shen,BAO Chang-chun |
| (Speech and Audio Signal Processing Lab, School of Electronic Information and Control Engineering,Beijing University of Technology, Beijing 100124, China) |
| Abstract: |
| According to the nonlinear dynamics theory, the phase space of MDCT（Modified Di
screte Cosine Transform） co
efficients of audio signals is reconstructed by computing delay time and embeddi
ng dimension in terms of autocorrelation method and false nearest neighbor (FNN)
method, respectively. The maximum Lyapunov exponents are computed by means of R
osenstein small data sets method. The chaotic characteristic of audio signals in
frequency domain is statistically analysed and verified based on the sign of th
e maximum Lyapunov exponents. The experiments show that the maximum Lyapunov exp
onents are almost positive for the MDCT series of audio signals,which demonstrat
es th
at the audio signals have the chaotic characteristics. |
| Key words: audio signal processing modified discrete cosine transform(MDCT) phase space r
econstruction Lyapunov exponent ch
aotic characteristic |
