Hybrid decision feedback equalization

Author

Jayaraman, Srikant ; Fernandez-Corbaton, Ivan ; Smee, John E.

Author_Institution

QUALCOMM, Inc., San Diego, CA, USA

Volume

4

fYear

2003

fDate

11-15 May 2003

Firstpage

2380

Abstract

By modeling the effects of slicer errors, we derive an algorithm for determining hybrid DFE coefficients which account for the increase in interference from error propagation. The coefficients of the hybrid DFE can be determined adaptively through the LMS algorithm by including an SINR dependent "leakage" factor in the feedback filter update equation. The hybrid DFE coefficients reduce to those of a linear equalizer (LE) at low SNR and a DFE at high SNR, while achieving an optimal compromise between these extremes. In the SNR region of greatest interest for systems with high spectral efficiency, simulation results show that the hybrid DFE offers significant gains compared to a regular DFE or an LE of comparable complexity.

Keywords

decision feedback equalisers; least mean squares methods; radiofrequency interference; DFE; LMS algorithm; SINR; comparable complexity; error propagation; feedback filter update equation; hybrid decision feedback equalization; interference; least mean square; linear equalizer; signal-to-inference-plus-noise ratio; slicer error effects; spectral efficiency; Constellation diagram; Data communication; Decision feedback equalizers; Decoding; Error analysis; Interference; Nonlinear filters; Signal to noise ratio; Transmitters; Turbo codes;

fLanguage

English

Publisher

ieee

Conference_Titel

Communications, 2003. ICC '03. IEEE International Conference on

Print_ISBN

0-7803-7802-4

Type

conf

DOI

10.1109/ICC.2003.1204340

Filename

1204340