Training-Assisted Channel Estimation For MIMO-OFDM Systems

Orthogonal frequency division multiplexing (OFDM) converts a frequency-selective channel into a parallel collection of frequency flat subchannels, thereby eliminating the effect caused by the frequency-selection fading. The OFDM may be combined with antenna arrays at the transmitter and receiver to increase the diversity gain and/or to enhance the system capacity. In this paper, we discuss the principle of MIMO-OFDM systems based on the introduction of fading channels. Furthermore, the channel estimation technique in MIMO-OFDM system is investigated, with the emphasis on training-assisted channel estimation methods. We first introduce a least squares (LS) channel estimation scheme for MIMO-OFDM systems based on training. After LS estimation, the mean square error (MSE) is computed. According to this MSE, the optimal training is derived. For this optimal training, a simplified implement based on lowpass filter is proposed. By simulation this implement is known as lower complexity than the conventional implement approximately same performance. Then some estimation schemes are discussed for some novel training patterns. By using the STBC training pattern, the complexity of MMSE channel estimation can significantly reduced and the performance can improved greatly in price of more training symbols. The number of the transmitter antenna being limited in the estimation schemes available, this paper proposed a antenna grouping method to meet the situation that the transmitter antenna are too many. This method is based on STBC training pattern. Lastly, the channel estimation in the data transmission mode is introduced. An expectation-maximization (EM)-based maximum-likelihood (ML) channel estimation algorithm is discussed in detail. By simulation, we know that the EM-based algorithm converges to the ML estimation within several iterations.