Advanced k-step forward prediction algorithm in wireless channel

In wireless transmission systems using adaptive techniques, a very important issue is the long-term estimation and prediction of Rayleigh channel fading. Considering the correlation character between sampled symbols and predicted ones on a time-varying channel, and the disadvantage of the normal k-step algorithm, an advanced novel k-step forward prediction algorithm, which is based on sub-predicted symbols, is proposed. The analysis shows there is an improvement in channel prediction and BER, even in a fast moving environment, by effectively exploiting more of the correlation information of channel samples than the normal k-step one does in a fast fading wireless channel. Simulations on advanced k-step also show a performance improvement, compared to the normal k-step, by lower sampling methods in a constant envelope system with a continuous power control scheme in order to further extend the predicting time.