A Novel Two-Dimensional Channel Estimation Algorithm for SC-FDE Systems

A novel time-frequency domain two-dimensional MMSE channel estimation (2D-MMSE) algorithm based on separation property of mobile wireless channels is proposed in this paper in order to provide the current channel state knowledge and improve the spectral efficiency in Single Carrier Frequency Domain Equalization (SC-FDE) systems over frequency-selective channels. We use the Unique Word (UW) to be the cyclic prefix and the pilot just like IEEE802.16. The UW is not need to delete in receiver, so the overhead of system is low. The 2D-MMSE estimation is simplified as an MMSE interpolation in the frequency domain cascaded by an MMSE filter in the time domain. The SVD method is adopt to reduce the complexity and keep the estimation accuracy. We also make the noise variance estimation by UW. Finally, the simulation results show that the proposed algorithm has better performance than the LS and MMSE algorithms, and the complexity of our method is lower than traditional one-dimensional MMSE.