Semi-blind Channel Estimation For Mimo Systems

Multi-antenna transmission tends to be an important technique for future wireless communication systems. Systems equipped with multiple transmit and receive antennas are referred to as multi-input and multi-output (MIMO) systems. In an environment rich of scattering, MIMO systems offer significant channel capacity gain over single-input single output (SISO) systems within the same spectrum.Due to the complex, unknown, and continuous valued MIMO channel coefficients, the work of receiver recovering the transmitted data as correct as possible becomes a challenge work. To achieve this, the semi-blind channel estimation (SBCE) has been widely employed due to the advantages of reasonable computational complexity, high accuracy and the capability to conserve the system bandwidth efficiency.In this thesis, the performances of SBCE with a-posteriori feedback and with extrinsic feedback in uncoded MIMO systems, coded SISO systems and coded MIMO systems was compared respectively. After that, the research work was focused on SBCE algorithms and a modified algorithm was proposed. The simulation results show that in an uncoded MIMO system, the breaking of noise error correlation and the error propagation enhances the performance of data detection and channel estimation. It can also compensate the decrease of the quality of channel estimation caused by the absence of one vector of useful information when increasing the number of receiver antennas. In a coded system, especially for a coded MIMO system, the algorithm with temporal extrinsic feedback presents that breaking the error propagation by re-estimate the data can improve the performance of the data detection.