Impact of imperfect channel knowledge on the performance of multicarrier systems

We discuss the effect of imperfect knowledge of the transmission channel on the performance of multicarrier-OFDM and DMT -systems. The limited accuracy in the identification of the transmission channel is caused either by errors in the estimate of the channel or by variations in the channel due to motion of the transmitter, the receiver, and/or the reflectors. We develop a model to analyze the impact of these errors and derive analytic expressions for the signal-to-interference ratio (SIR) and the probability of error as a function of the initial SNR and the mean square error in channel estimation. We compare these results to those obtained assuming Gaussian errors and show that in both cases, the performance of multicarrier systems degrades when the normalized mean square error in channel estimation is greater than 10/sup -3/ (-30 dB). Finally, we extend these concepts to errors due to variations in the transmission channel.