分数阶傅立叶变换用于信号分析的研究信息与通信工程专业论文.docx

哈尔滨工业大学工学硕士学位论文 哈尔滨工业大学工学硕士学位论文 - - II - Abstract The fractional Fourier transform (FrFT), as the generalization of conventional Fourier transformation, is actually a new transform, which reflects the characteristics of a signal in the time-frequency domain. The conception of FrFT was firstly proposed by Namias in 1980. In the following two decades, researchers made lots of efforts to build the theory of FrFT. The physical meaning and discrete algorithm and other achievements of FrFT theory made it more than a mathematics theory .FrFT became an new tool for signal analysis. In recent 10 years, lots of articles focused on signal processing in fractional Fourier domain were published The dissertation starts with the definition, property and the numeric compute method of FrFT and carries out numeric compute imitation of the desirable properties of the FrFT and demonstrates its relationships with some well-known signal transforms such as the Wigner–Wille Distribution (WVD). In order to apply FrFT to signal analysis, the convolution and correlation algorithm were studied. Then the text analyzed the ordinary signals in traditional communication systems in Fractional Fourier domain. The results of imitation show us the ability of FrFT to reflect the characteristics of a signal in the time-frequency domain. Founded on the previous research, FrFT has ability to analyze and process non-stationary signals. Meanwhile, FrFT has a chirp element in the kernel function, so it is propitious to deal with Linear Frequency-Modulated signal. The dissertation made numeric compute imitation of Detection and Parameter Estimation of single and Multi-component LFM Signal With a self-adaptive filter structure Based on FrFT. The results showed that FrFT has excellent ability to analyze and process LFM-like non-stationary signals. In the end, the dissertation studied the structures of Digital communication system based on LFM modulation and demodulation. Keywords FrFT; signal analysis; LFM; -