Title :
Hybrid HOS-SOS approach for blind equalisation of communication channels
Author :
Hussain, A. ; Naveed, A. ; Qureshi, I.M.
Author_Institution :
Centre for Cognitive & Comput. Neurosci., Univ. of Stirling, UK
fDate :
3/17/2005 12:00:00 AM
Abstract :
A new hybrid higher-order statistics (HOS) and second-order statistics (SOS) based approach to improve the performance of the standard Bussgang algorithm for blind equalisation of digital communication channels is presented. An additional term, based on SOS, has been added to the conventional cost function, and the Bayesian estimator has been replaced with an adaptive activation function. This is shown to give faster convergence by an order of magnitude, both in terms of epochs and the mean square error between input and output symbols at different signal-to-noise ratios (SNRs).
Keywords :
blind equalisers; convergence of numerical methods; digital communication; higher order statistics; mean square error methods; statistical analysis; telecommunication channels; Bayesian estimator; Bussgang algorithm; HOS; SNR; adaptive activation function; blind equalisation; convergence; cost function; digital communication channel equalisation; higher-order statistics; hybrid statistical approach; mean square error; second-order statistics; signal-to-noise ratio;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:20057272
