Title :
A neural network for smallest eigenvalue with application to blind equalization
Author :
Dong, Guojie ; Liu, Ruey-wen
Author_Institution :
Dept. of Electr. Eng., Notre Dame Univ., IN, USA
Abstract :
In this paper, a neural network is presented which can extract the smallest eigenvalue and its associated eigenvector of the autocorrelation matrix of a incoming vector of stochastic process. It is also shown that the domain of convergence is the unit sphere, substantiated by computer simulation. We also demonstrate by simulation that this neural network is capable to do blind equalization which is crucial for wireless communication
Keywords :
convergence; correlation theory; eigenvalues and eigenfunctions; equalisers; neural nets; stochastic processes; autocorrelation matrix; blind equalization; computer simulation; convergence; eigenvalue; eigenvector; neural network; stochastic process; vector; wireless communication; Application software; Autocorrelation; Blind equalizers; Computational modeling; Computer simulation; Convergence; Eigenvalues and eigenfunctions; Neural networks; Stochastic processes; Wireless communication;
Conference_Titel :
Circuits and Systems, 1995., Proceedings., Proceedings of the 38th Midwest Symposium on
Conference_Location :
Rio de Janeiro
Print_ISBN :
0-7803-2972-4
DOI :
10.1109/MWSCAS.1995.510212
