Channel Estimation for Macrocellular OFDM Uplinks in Time-Varying Channels

This paper deals with channel estimation in time-varying channels for orthogonal frequency division multiplexing (OFDM) uplink transmission. By modeling the uplink channel properly, the time-varying channel response can be determined by estimating the channel parameters. These unknown channel parameters are coupled with each other due to multipath propagation. By exploiting the orthogonality of the training symbol, the channel parameters can be easily separated. Therefore, we can estimate the unknown channel parameters path by path, instead of complex joint estimation. This leads to great reduction of computational complexity. Moreover, an order-recursive algorithm is proposed, which can approximate the nonlinear nature of Doppler shifts with any-order Taylor expansion rather than only the second-order one in the literature. The proposed algorithm can outperform the existing algorithms due to the employment of higher order Taylor expansion. Theoretical analysis and simulations are also given, demonstrating the efficiency of the proposed algorithm.