Research On Channel Estimation Algorithms In MIMO-OFDM Wireless Communication Systems

Multiple-Input Multiple-Output (MIMO) and Orthogonal frequency division multiplexing (OFDM) technology have the advantage of improving the spectral efficiency and resistance to multipath fading respectively. MIMO-OFDM combine these two techniques and improve the channel capacity and transmission rate of wireless communication systems greatly, which is recognized as one of the core technology of next generation wireless communication systems. Channel estimation is a key factor in MIMO-OFDM systems for coherent detection and accurate decoding.This paper takes channel estimation of MIMO-OFDM system as the main content of the study. For the accuracy of the existing estimation algorithm is not high with the effect of inter-carrier interference (ICI) influence, this paper give the improved algorithm.First, for the low precision of interpolation algorithm after the least squares method, this paper proposes Hermite interpolation algorithm for channel interpolation. Compare with linear interpolation and transform domain interpolation algorithm, Hermite interpolation algorithm does not only consider the curve trend, but also consider the slope of the interpolation point, and make the interpolation more accurate with low complexity.Second, for the iterative least square (RLS) does not consider the ICI effect, this paper introduce frequency-dependent encoding and decoding to the transmitter and the receiver, the introduction of correlation between adjacent carrier can suppress the interference between the sub-carrier in MIMO-OFDM system, which can improve the performance of MIMO-OFDM systems effectively.Third, for the larger Doppler frequency shift and delay, this paper take the piece-wise linear approximation to the channel estimation and improve it. This paper proposed the piece-wise linear approximation method to reach the channel ICI cancellation within an OFDM symbol, then eliminating the ICI and estimating the channel response accurately. Put a training OFDM symbol at first, and determine the value of the slope of the first OFDM symbol, and then estimate the channel response accurately.Finally, this paper take three improvement algorithms above for simulation, simulation shows that the channel signal carrier interference power ratio(CIR), the channel estimation mean square error(MSE), the channel estimation error rate(BER) have been improved greatly after the ICI cancellation.