Etude et Simulation de la Prédiction d’un Canal Radio-Mobile par le Processus Autorégressif

Abstract

The propagation channel prediction is a fundamental technique in Wireless communications systems, in particular in case of mobility where the channel parameters vary with time. Prediction allows, in this case, to know the future values of complex channel gains from past values. Among the techniques proposed in the littérature, the one which is based on the autoregressive process (AR), generally used for the prediction and estimation of time series, was chosen for this work. The performance of this technique with point of view of symbol error rate versus signal to noise ratio has been studied and analyzed in the case of single carrier and multi-carrier transmission with variable speed where the propagation channel is Rayleigh Jakes type. Key-words : AR, Doppler frequency, Rayleigh, OFDM, Yule-Walker, MSE.