Title :
Identifiability of blind deconvolution of non-white signals
Author_Institution :
Dept. of Electr. & Comput. Eng., Drexel Univ., Philadelphia, PA, USA
Abstract :
Considers the problem where an unobservable, non-Gaussian and non-white stochastic information bearing signal is transmitted through two different linear time invariant systems (LTI), and there are only noisy observations of the system outputs. The goal is to identify and deconvolve the signal within a scalar and a time delay. Apparently this type of problem is ill-posed. Identifiability of the information bearing signal is defined with respect to a set of conditions that specify some of the properties of the signal and the LTI systems. Such sets of conditions are defined, and deconvolution schemes based on these sets are studied.
Keywords :
"Deconvolution","Signal processing","Stochastic systems","Time invariant systems","Delay effects"
Conference_Titel :
Signals, Systems and Computers, 1993. 1993 Conference Record of The Twenty-Seventh Asilomar Conference on
Print_ISBN :
0-8186-4120-7
DOI :
10.1109/ACSSC.1993.342373