Study And Realization Of Semi-blind Channel Estimation Algorithm In MIMO-OFDM System

As one of the key technologies of the fourth generation (4G) mobile communication systems, MIMO-OFDM technology has been widely applied in communication signal processing in recent years. MIMO-OFDM is the technology combining space, frequency and time diversities, which can not only greatly improve tansimission rate and channel capacity of wireless communication, but also efficiently combat multi-path fading and interference.Based on the analysis of the fading characteristic of mobile wireless channel, this paper mainly studies the semi-blind channel estimation technique in MIMO-OFDM system. A blind channel estimation algorithm, subspace method, has been analysed in the literature. The method has many problems such as uncertainty of estimation. In order to overcome the problems, a semi-blind estimation algorithm which combines subspace method with the sounding-based channel estimation algorithm is given. Using Clarke model of Rayleigh channel to simulate, it is indicated that the performance of the algorithm is fine, but its computation is too high to guarantee a real-time processing. For this problem, using the high correlation characteristic of the adjacent sub-carriers, this paper proposes an improved semi-blind algorithm, which not only improves the performance, speed up the convergence rate, but also greatly reduces the computing complexity to guarantee a real-time processing.Finally, the improved method is realized by FPGA using three modules. The overall timing and resource consumption are also analyzed.