Study On Channel Estimation Method For Wireless OFDM Systems

Orthogonal Frequency Division Multiplexing (OFDM) is a multi-carrier modulation technique. For the reason that it has strong ability to defend multi-path effect and high bandwidth efficiency, so it has become emphasized more and more. Especially, OFDM has shown great performance in high data transmission, it will become the first scheme for 4G technology of wireless communication in the future.Generally, the demodulation way can be classified as differential demodulation and non-differential demodulation. Meanwhile, non-differential demodulation can be classified as coherent demodulation and non-coherent demodulation. The differential demodulation doesn't need the known channel state information, and it only suits to the case that the data transmitted in low rate. On the contrary, coherent demodulation can be applied in high data transmission and it needs the known channel state information. In order to meet the need of high data transmission, the coherent modulation way is often adapted in practical OFDM systems. Because that wireless channel is unknown in practice, it is necessary to estimate the channel's state information. The methods of iteration to estimate channel's state information in the WSSUS channel environment are introduced as an importance in this paper. Based on it, a windowed linear prediction method under the MMSE law is presented for estimating the channel initial frequency response. For the reason that the method makes full use of the statistical property and the correlation of channel frequency response, therefore, the more precise estimation value can be obtained. Simulation results indicate, although the performance of systems employing estimation method of adding window linear prediction based on MMSE law is almost as same as that of other methods in low SNR, it is better than that of other algorithms in high SNR region.The method is an effectively improved algorithm with moderately increased complexity.