Decision feedback channel estimation of OFDM systems in fast time-varying channels

In order to improve the transmission performance of Orthogonal Frequency Division Multiplexing (OFDM) systems in fast time-varying channels, a novel time-domain channel estimation scheme based on the superimposed training sequence is proposed. In this scheme, according to the exponential interpolation model of the channel tap coefficient, a linear model is established by using the periodic superimposed training sequence of a symbol, and the parameters of the interpolation model are estimated by means of the least square method. Then, the corresponding time-domain response matrix and frequency-domain equalization matrix are constructed. Moreover, in order to further improve the estimation and equalization performances, symbol decisions are taken as the known training sequences to estimate the iterative channel parameters, thus effectively improving the channel estimation performance and eliminating the inter-carrier interference. Experimental results indicate that the proposed scheme is of excellent performance in the condition of time-frequency doubly-selective channels.