An analytical method of channel estimation error for multiple antennas system

In the flat-fading channel (narrow bandwidth channel), the channel coefficients are generally obtained by sending training sequences or pilot signals in the transmitter and estimating them in the receiver. Therefore, some estimation errors are induced in this process. In the paper, we construct a kind of model of channel coefficient estimation error according to the general theory of multiple antennas wireless link, and on the basis of analyzing the statistical characteristics of channel estimation error, and we derive the Cramer-Rao bound of the above estimation error by virtue of the model. It is found that the Cramer-Rao bound given in (V. Tarokh, et al., 2000) is a special case of our result given in the paper. As the ratio of signal to noise is large enough, the former is derived from the latter. Space-time block coding (STBC) is taken as an example, and the loss of system performance is simulated in the presence of channel coefficient estimation error. The simulation results show that the loss of system performance of STBC caused by the channel coefficient estimation error is less than 1 dB, and within the Cramer-Rao bound of estimation error, less than 0.5 dB, which is the minimum in the losses of system performance caused by all of the channel coefficient estimation errors. This conclusion is in accord with the analytical result in theory.