Normalized step size approach to signal processing based on lagged cross-correlation of probability

초록

The signal processing algorithms based on the lagged cross-correlation of probability (LCCP) have been considered very effective in the multipaThenvironment with impulsive and DC noise coexisting in visual light/power line communication (VLC/PLC) systems for energy saving intelligent buildings. To meet the demand for higher data rates and more complicated sensor node configuration, in this paper, performance enhancement methods by employing a normalized step size in the LCCP-based algorithms which are sensitive to the scaling of system input contaminated with impulses and DC noise are proposed through the analysis of the behavior of optimum weight and the role of magnitude controlled input (MCI) on influence of large errors. The normalized LCCP algorithm with the step size normalized with the averaged power of the current MCI shows a faster speed by about 5 times and has a lower minimum mean squared error by about 3 dB than the conventional LCCP algorithm in blind equalization simulation for indoor VLC/PLC channel models. © 2005 - 2016 JATIT & LLS. All rights reserved.

키워드