Title :
Digital satellite channel identification using neural networks: analytic models
Author :
Bershad, N.J. ; Ibnkhala, M. ; Castanie, F.
Author_Institution :
Dept. of Electr. & Comput. Eng., California Univ., Irvine, CA, USA
Abstract :
This paper studies the statistical transient and convergence behavior of a neural network structure (filter-nonlinearity-filter) that adapts its parameters using a modified version of the backpropagation algorithm. A previous paper has used this structure to model and identify nonlinear satellite channels with memory. This paper derives simplified analytical models which qualitatively predict the algorithm behavior observed previously.
Keywords :
adaptive estimation; backpropagation; convergence of numerical methods; digital communication; filtering theory; identification; neural nets; satellite communication; statistical analysis; telecommunication channels; analytic models; backpropagation algorithm; digital satellite channel identification; filter-nonlinearity-filter; memory; neural networks; nonlinear satellite channels; statistical convergence behavior; statistical transient behavior; Adaptive filters; Backpropagation algorithms; Convergence; Delay effects; Neural networks; Nonlinear filters; Predictive models; Satellites; Signal processing algorithms; Transversal filters;
Conference_Titel :
Signal Processing Advances in Wireless Communications, First IEEE Signal Processing Workshop on
Conference_Location :
Paris, France
Print_ISBN :
0-7803-3944-4
DOI :
10.1109/SPAWC.1997.630378
