J4

• Original Articles • Previous Articles     Next Articles

A method for ICA for complex-valued sources

LI Xiao-jun1;LOU Shun-tian1;ZHANG Xian-da2

  

  1. (1. State Key Lab. of Radar Signal Processing, Xidian Univ., Xi′an 710071, China ;
    2. Dept. of Automation, Tsinghua Univ., Beijing 100084, China)
  • Received:1900-01-01 Revised:1900-01-01 Online:2005-06-20 Published:2005-06-20

Abstract: ICA is used to get the complex-valued signals. ICA is a statistical method for transforming an observed random vector into components that are as mutually independent as possible. By changing the extended Hebbian learning and directly using the nonline function, blind separation of complex-valued signals will be obtained. By using this method, the algorithm based on the gradient steepest ascent is proposed, with its computational efficiency shown by simulations.

Key words: ICA, Hebbian learning, gradient steepest ascent

CLC Number: 

  • TN911.23