Improved power efficiency of the LC-LMS equalizer through partial elimination of the constraint update

We investigate the conditions for which elimination of the square root raised-cosine matched filter processing of the linearly constrained LMS equalizer leads to reduced computational complexity and improved power efficiency of this constrained equalizer. We demonstrate that the extent of the power efficiency is dependent upon the value of the adaptation constant associated with channel inversion. Previous literature (F. Haris et al., (2002) which stressed the ability of this joint process equalizer to conserve the hardware of the square root raised-cosine matched filter, combined with the potential for power efficiency emphasized in this paper, promote a new class of fractional-spaced equalizers that efficiently utilize 100% of the spectral band to reduce the cost of digital receiver implementation.