Title :
Discrete signal reconstruction from its spectral magnitude and some samples
Author :
Wu, Zhongze ; Li, Yenta ; Chang, Tong
Author_Institution :
Tsinghua University, Beijing, China
Abstract :
Discrete signal reconstruction from its spectral magnitude and some samples has a variety of potential applications. In this paper, a discrete signal is shown to be uniquely reconstructed by its spectral magnitude and some saples under mild restriction on its original sequence. Three theorems are provided. Several numerical algorithms have been discussed and a new algorithm called three domains iteration method in which the convergence is guaranteed is developed.
Keywords :
Automation; Fourier transforms; Prosthetics; Signal processing algorithms; Signal reconstruction; Speech processing;
Conference_Titel :
Acoustics, Speech, and Signal Processing, IEEE International Conference on ICASSP '86.
DOI :
10.1109/ICASSP.1986.1168939