Channel Estimation And Equalization For MIMO-OFDM Systems

MIMO-OFDM system can resist channel's frequency selective fading, enhance the system's capacity and have a good spectrum utilization. But it needs the channel state information to realize this, so channel estimation and equalization play an important role in the whole system. In this paper, these two key technologies for MIMO-OFDM systems are studied.Firstly, due to high computational complexity of mostly channel estimation methods, this paper presents a reduced-complexity minimum mean square error (MMSE) channel estimator for MIMO-OFDM systems, which is based on the expectation maximization (EM) algorithm by low-rank approximation to avoid inverting large-size matrices and by the use of statistical properties of channel correlation matrix to decrease frequently multiply operation. Simulation results show the proposed algorithm drastically reduces the receiver complexity and has comparable performance with the original MMSE algorithm.Secondly, signal transported in the fast fading channel will cause serious ICI. To eliminate subcarriers interference, an iterative algorithm in frequency domain is proposed. In the algorithm, channel transmission matrix is divided into two parts, namely data matrix and ICI matrix. The interference between adjacent subcarriers, created by ICI matrix, is then subtracted from received symbols, and finally the parallel interference cancellation detection and data matrix is performed to suppress the multi-stream interference from different antennas. Only considering the interference of neighboring subcarriers, under this condition the performance loss is neglectable compared with the original method, and computational complexity can be reduced. Simulation results show the proposed method can effectively mitigate the effect of ICI at different Doppler shifts and approach the ICI-free performance at low signal-to-noise ratio.