Optimal pilot design for MIMO OFDM channel estimation

It is well known that the realization of enormous capacity of multiple-input multiple-output (MIMO) channels depends greatly on the accuracy of channel estimation at the receiver. And the problem of pilot or training design is crucial to channel estimation. In this paper, the optimal pilot sequences and the optimal length of pilot-OFDM symbols are discussed. The optimizing process of pilot design consists of two phases. First, by minimizing the channel estimation error, the initial optimal pilot sequences are designed to exploit full information of channel correlation under the criterion of Minimum Normalized Mean Square Error (MNMSE) of the frequency domain, and at the same time, the optimal length of pilot-OFDM symbols for various range of SNR is also concluded. Secondly, in view of restraining the peak-to-average energy ratio (PAR) of the pilot-OFDM signal, the final optimal pilot sequences are obtained by choosing a unit complex-value sequence with lower aperiodic autocorrelation and using the initial optimal pilot sequences. Simulations show that the channel correlation is helpful to decrease the MNMSE of channel estimation and the designed optimal pilot sequences present great superiority in PAR-constraining.