Abstract
Multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) is a new transmission scheme that combines MIMO signal processing and OFDM techniques. In this thesis, we propose a new training sequence structure along with a complete synchronization scheme for MIMO-OFDM systems. The proposed synchronization scheme includes coarse time acquisition, coarse carrier frequency offset estimation, fine time acquisition, and fine carrier frequency offset estimation. In the proposed synchronization scheme, we adopt existing techniques to obtain initial timing and frequency offset estimates for the proposed fine time acquisition and fine frequency offset estimation methods. The proposed fine time acquisition method is assisted by the estimated channel gain, which is obtained by a recently proposed DFT-based channel estimator. The proposed approach is able to provide each transmit-receive branch with an accurate and dedicated timing instant. Based on the timing instants obtained, the proposed fine carrier frequency offset estimator decides a proper correlation range to obtain a better estimate. Computer simulation results show that, the proposed fine time acquisition method can provide significant performance gain over a conventional method. Unlike the conventional method, the proposed approach provides accurate and comparable estimates regardless of the number of transmit antennas. In addition, the proposed fine carrier frequency offset estimator can improve the coarse frequency offset estimator by finding a proper correlation range and averaging estimates obtained from different transmit-receive branches.