Study On Channel Estimation Algorithms For Mimo-OFDM System In Wireless Fading Channels

A multiple-input multiple-output and orthogonal frequency division multi–plexing(MIMO-OFDM) system is a system that employs multiple transmit andreceive antennas in an orthogonal frequency division multiplexing(OFDM)communication system. Using multiple transmitter and receiver can enhancesystem capacity, while using orthogonal frequency division multi-plexingtechnique can combat the multi-path delay spread and mitigate inter-symbolinterference.However, for the wireless mobile communication systems, the radiochannel is frequency selective and time-varying. A dynamic estimation ofchannel is necessary for diversity combining, coherent detection, and decoding.So, to improve the performance of MIMO-OFDM systems by accurate channelestimation, we will mostly investigate the channel estimation techniques ofMIMO-OFDM system. Based on the above analysis, this dissertation undertakessome research work on channel estimation and optimal pilot strategies based onthe system design.At first, a optimal pilot strategy in the transmitter diversity based on theMSE of system SNR was deduced, compared with the existing alternativetransmitting pilot methods in 802.16, the proposed pilots' algorithm can bescattered in all the pilot sub-carriers. Especially in mobile channels, theproposed algorithm shows lower MSE.Then, we propose an improved robust frequency domain diversity timecombining(FDTC) channel estimation algorithm. And the FDTC algorithmcan farther improve the performance, and have better robust and lowercalculating complexity. When the bias of synchronization exists, the FDTCalgorithm can revise the error of the phase offset. Comparing with thetrans-domain methods, the FDTC algorithm can avoid the serious aberrance,which is chiefly caused by the energy leakage induced by the zero bandscorresponding to the rectangular windowing.Finally, the analysis and comparison of these algorithms is made bySystemView communication simulation tool.