A Tap Weight Selection Method for Equalization of Wireless Channels with Large Multipath Delay

In this work we exploit the sparse nature of digital radio channels affected by large multipath delays towards complexity reduction of adaptive equalization algorithms. A sparse channel is defined here as the one having a few nonzero powerful taps separated by many negligible taps. The terrestrial High Definition Television (HDTV) channel model is considered here as a practical example of a sparse wireless channel. A tap-weight selection method is formulated based on the magnitude relationship between the negligible and the powerful tap weights of a linear equalizer. The proposed method eliminates those tap weights that are considered negligible and leave only the most powerful ones. Compared to a conventional adaptive equalizer its possible to achieve at least the same performance using only the most important (powerful) tap weights. Furthermore, the tap-weight selection (TWS) method results in a considerable computational resource saving, which is proportional to the number of negligible taps of the linear equalizer.