A NOVEL BLIND EQUALIZATION STRUCTURE BASED ON CMA-GENERATED DESIRED RESPONSE
2000
Hochschulschrift
Zugriff:
88
It is well known that conventional constant modulus algorithm (CMA)-based blind equalizers exhibit slow rate of convergence and large steady state mean square error (MSE). In this thesis we propose a novel structure for blind equalization that includes a blind part and a training-like part. In the blind part, one of CMA-based algorithms such as the standard CMA is employed to generate an estimate of the desired data symbol that is used as the desired response by the training-like part. In the training-like part whose function is essentially the same as that of the standard trained equalization, one of the conventional training algorithms such as the recursive least-squares (RLS) algorithm is employed to improve the convergence process as well as the steady state MSE. Computer simulations demonstrate that the proposed structure that is applicable to both constant and non-constant modulus signals can achieve excellent performance in rate of convergence as well as steady state MSE for blind equalization.
Titel: |
A NOVEL BLIND EQUALIZATION STRUCTURE BASED ON CMA-GENERATED DESIRED RESPONSE
|
---|---|
Autor/in / Beteiligte Person: | Liao, Yi-Cheng ; 廖益正 |
Link: | |
Veröffentlichung: | 2000 |
Medientyp: | Hochschulschrift |
Sonstiges: |
|