Title :
Minor component analysis with implementation to blind 2-channel equalization
Author :
Luo, Jie ; Ling, Xieting
Author_Institution :
Fudan Univ., Shanghai, China
Abstract :
A null space method is proposed for blind identification of non-minimum phase linear time-variant channels using minor component analysis. The algorithm is proved to be globally convergent in discrete domain. The fast convergence speed makes this algorithm very useful for tracing a time-variant channel in mobile communication. Some simulations are given to show its tracking ability, and a very efficient method for equalizing the blind channel is also proposed
Keywords :
convergence of numerical methods; eigenvalues and eigenfunctions; equalisers; identification; mobile communication; neural nets; statistical analysis; time-varying channels; blind channel; blind equalization; blind identification; discrete domain; eigenvalues; globally convergence; minor component analysis; mobile communication; neural network; nonminimum phase channels; null space learning; time-variant channel; Blind equalizers; Communication channels; Computer simulation; Convergence; Eigenvalues and eigenfunctions; Mobile communication; Neural networks; Null space; Signal processing; Statistics;
Conference_Titel :
Neural Networks,1997., International Conference on
Conference_Location :
Houston, TX
Print_ISBN :
0-7803-4122-8
DOI :
10.1109/ICNN.1997.614375
