Proceedings of 1994 28th Asilomar Conference on Signals, Systems and Computers (1994)
Pacific Grove, CA, USA
Oct. 31, 1994 to Nov. 2, 1994
R.H. Lambert , Dept. of Electr. Eng. Syst., Univ. of Southern California, Los Angeles, CA, USA
C.L. Nikias , Dept. of Electr. Eng. Syst., Univ. of Southern California, Los Angeles, CA, USA
A new technique for blind equalization is presented which estimates the inverse (deconvolution) multipath channel and the forward channel as well. This method removes the need for an ad hoc energy constraint on the filter taps and constrains the equalization filter by relating the output back to the equalizer input. The concept of feedback is used to provide a more meaningful constraint for blind equalization filters. The real advantage of the method is increased speed of convergence over traditional stochastic gradient methods, and greater robustness to eigenvalue spread. These benefits come at a modest two times the computational complexity price.<
computational complexity, equalisers, feedback, deconvolution, multipath channels, eigenvalues and eigenfunctions, convergence of numerical methods, filtering theory
R. Lambert and C. Nikias, "Forward/inverse blind equalization," Proceedings of 1994 28th Asilomar Conference on Signals, Systems and Computers(ACSSC), Pacific Grove, CA, USA, 1995, pp. 1463-1467.