Efficient channel estimation for MIMO single-carrier block transmission with dual cyclic timeslot structure

Abstract—We investigate channel estimation for timeslot-structured single-carrier block transmission (SCBT) over space-, time-, and frequency-selective fading multiple-input multiple-output (MIMO) channels. A MIMO-SCBT with a dual cyclic timeslot structure is presented first. Then, an optimal chan-nel estimation in the minimal mean square error (MMSE) sense on the timeslot basis is investigated. It is shown that the optimal pilots for the timeslot-based MMSE channel estimation are related to the statistical channel state information in eigenmode. Under the assumption that the transmit correlation is unknown at the transmitter, the optimal pilots satisfy the same condition as re-ported for the block-based least-square (LS) channel estimation in literature, and the channel estimation can be simplified to initial block-based LS channel estimation followed by space–time post-processing. Particularly, for spatially uncorrelated channels, the space–time postprocessing can be reduced to pathwise processing. A new design of the pilot sequences is given, which leads to an effi-cient implementation of the channel estimation. Later on, a more efficient implementation for the initial channel estimation is ob-tained by using the structure of the pilot sequences, and discrete cosine transform (DCT)-based implementation is developed for the space–time postprocessing to approximate the optimal solution with low implementation complexity. Finally, the performance of the proposed channel estimation is verified via simulations. Index Terms—Channel estimation, multiple-input multiple-output (MIMO), single-carrier block transmission (SCBT). I