A recently proposed simple transmit diversity (STD) scheme was shown to hav
e the same performance as the classical maximal-ratio combining (MRC) recei
ver under the assumption that the amplitude and phase of the Rayleigh faded
channel are known exactly at the receiver. An analysis and comparison of t
he performance degradations of the STD scheme and MRC receiver due to error
s in estimating the channel parameters are presented. It is shown that the
bit error rate degradation for the STD scheme can be significantly worse th
an that of an MRC receiver.