Subspace-based Semi-blind Channel Estimation For MIMO-OFDM Systems

Multi input multiple output(MIMO) technology can make full use of the multipath effect of the wireless channel, and can greatly improve the system capacity without increasing the system bandwidth. The orthogonal frequency division multiplexing(OFDM) technology can effectively inhibit the fading of the wireless channel and improve the spectrum utilization. So, the combination of the MIMO- OFDM technology has become a hot research direction. In the MIMO-OFDM system, accurate channel estimation is the guarantee for the high-speed data transmission and high reliability. The spectral efficiency of the non- blind estimation system is low, while the complexity of the blind estimation is high and the estimated value is imprecise. Semi blind estimation is to combine the two, in blind channel estimation based on adding a few pilot symbols, which better solves the estimated value of the uncertain problems, improve the utilization rate of the spectrum.The main work of this paper is:(1) The method of reducing rank orthogonal iteration is discussed. By transforming the self-correlation matrix into another matrix, and using the QR decomposition to replace the traditional SVD method to the noise subspace semi blind channel estimation, thus reducing the computational complexity. Finally, the theoretical analysis and simulation results show that: the proposed method can reduce the computational complexity of the channel estimation and have good convergence effect in the case of the security.(2) Based on FRANS tracking method and the iterative tracking used to the noise subspace estimation to. Finally, the eigen decomposition of the resultant matrix, to realize the system of semi blind channel estimation is discussed. By the theoretical analysis and simulation results show that, the method to reduce the computational complexity at the same time in channel estimation, can also change the learning factor value to improve the error performance of the system, but there are also shortcomings of large amount of calculation.