Research On The Semi-Blind Channel Estimation In MIMO-OFDM Systems

Compared with the 3G mobile communication system, the B3G system has a higher requirement for the data rate and utilization efficiency of the spectrum. For high-speed mobile communication system the channel fading is more serious than ever. The MIMO (Multi-Input Multi-Output) system can solve the problem of fading and substantially increase the channel capacity. On the other hand, OFDM (Orthogonal Frequency Division Multiplexing) can efficiently combat the channel frequent selective fading. So, OFDM and MIMO with STC (Space Time Coding) becomes the kernel technology of the B3G system.Decoding of STC requires the channel state information at the receiver and the accuracy of the channel estimation determines the performance of a communication system. So channel estimation becomes a key technology in MIMO-OFDM system. Pilot assisted estimation and blind estimation are two well researched methods. Blind estimation has the advantages of high BER (Bit Error Rate) and data transition efficiency. But it is more complex to implement and has a slow rate of convergence. Pilot assisted estimation is much simpler in implementation and attracted much attention on it. But it is not so efficient in data transition and weak in channel tracking and things get worse when system requires high data rate and high mobility. As the result, semi-blind channel estimation which has both of the advantages becomes the hot topic. The dissertation will research on the semi-blind channel estimation in MIMO-OFDM system. The dissertation first introduced the history of mobile communication and the features of the B3G system. After that, the basic knowledge of wireless channel was introduced, and then the system module of MIMO-OFDM.Two common methods of channel estimation (pilot assisted and blind) were explained, and conclusions were made based on the analysis of their principles and features.Then the dissertation raised two algorithms of semi-blind channel estimation, based on adaptive filters and subspace respectively. The first algorithm used Kalman filters and time domain pilot to estimate and track the variant channel. It has a high data transition rate and good tracking ability. The second algorithm used subspace concept together with STC-OFDM to solve the problem caused by traditional subspace method. Then the dissertation proposed a method, which used frequency domain pilot assisted estimation, to mitigate the interference caused by noise and multi-antenna. Simulation results indicated that the two algorithms can achieve BER of 10?3 in high SNR condition and that the band usage for pilots is very small. At last the dissertation pointed out that although it is relatively complex in implementation, semi-blind channel estimation method should be very promising as the development of the hardware.