Title :
Quasi-minimum-BER linear combiner equalizers
Author :
Tellado, J. ; Cioffi, John M.
Author_Institution :
Inf. Syst. Lab., Stanford Univ., CA, USA
Abstract :
It is well known that the minimum mean square error (MMSE) and minimum peak distortion (MPD) criteria do not minimize bit error rate (MinBER), and for some channels they can have a large degradation compared to the MinBER solution. Based on the strengths and weaknesses of the MMSE and MPD criteria, we have defined two new solutions, which we call minimum noise limited peak distortion (MNLPD) and min-α. The first solution is based on MPD but avoids the problem of noise enhancement. The second solution finds the optimal of a family of equalizers derived from the MMSE and MPD criteria. We have shown by numerical calculation that by using the MNLPD or the min-α criteria we can achieve almost MinBER performance. The proposed methods have a larger complexity, but can be solved using well known techniques since they can be described as simple convex optimization problems
Keywords :
computational complexity; decision feedback equalisers; error statistics; least mean squares methods; minimisation; noise; MMSE; MNLPD; bit error rate; complexity; convex optimization problems; min-α; minimisation; minimum BER performance; minimum mean square error; minimum noise limited peak distortion; minimum peak distortion; optimal equalizers; quasi-minimum-BER linear combiner equalizers; Bit error rate; Decision feedback equalizers; Delay estimation; Digital communication; Equations; Error analysis; Filters; Mean square error methods; Vectors;
Conference_Titel :
Communications, 1998. ICC 98. Conference Record. 1998 IEEE International Conference on
Conference_Location :
Atlanta, GA
Print_ISBN :
0-7803-4788-9
DOI :
10.1109/ICC.1998.682535