Research On Dft-based Channel Estimation Algorithms For Mimo-ofdm Systems

Channel state information is required both in the coherent demodulation of OFDM systems and in the space-time decoding of MIMO systems. As one of the key technology of future wireless communication, channel estimation will play an important role in the whole system. Among all of the algorithms, the DFT-based channel estimation could achieve higher accuracy at low complexity. Hence, in MIMO-OFDM systems, the DFT-based channel estimation methods are investigated, and two improved algorithms are proposed in this paper.The first algorithm utilizes the property of the channel power which is more concentrated in the time domain. The significant channel taps are selected on the basis of an optimal threshold, which is derived in multiple antennas systems. The channel samples, which are larger than the optimal threshold, contain almost all the channel information. While the ignored channel samples contain noise energy only. Therefore the channel estimation algorithm based on the optimal threshold reduces the effect of the noise greatly. Each multipath parameter is characterized by a time delay and complex amplitude. Unlike the fast temporal variation of the complex amplitudes, the multipath delays are stationary over long time intervals. In order to acquire the multipath delays, the second algorithm takes advantage of a polyphase codes with good periodic correlation properties. According to the sparsity of the multi-path channel, the channel estimation based on the information of delays decides the significant channel taps by estimating the multipath delays and can also remove the effect of noise.In order to improve the accuracy, the core idea of the proposed methods is deciding the significant channel taps to restrain most of the noise. Without the channel statistics, both algorithms are less complicated and more practical. Both methods are simulated in 3GPP channel, and the validity of the proposed algorithms is verified.