Research Of Channel Estimation And Adaptive Modulation Algorithms In MIMO/OFDM Systems

Channel state information is required in OFDM systems for correlation detection and in MIMO-OFDM systems for space-time decoding. Meanwhile, it is the required information to the transmitter in closed-loop systems. Therefore, channel estimation is very important, especially, its accuracy directly affects the overall performance of MIMO-OFDM systems. On the other hand, adaptive modulation technique can enhance data rate, spectrum efficiency and transmission reliability in MIMO-OFDM systems. This dissertation is concerned with channel estimation in both OFDM and MIMO-OFDM wireless communication systems and adaptive modulation in MIMO-OFDM systems. The main contributions of this dissertation are recapitulated as follows. Firstly, a radial-basis-function-network-based channel estimation and tracking algorithm is presented for the OFDM systems with scattered-type pilots. This algorithm adopts grouped combination channel estimation and tracking methods based on cyclic shift property of the pilots in order to utilize both time and frequency correlation of channel response. The simulation and analysis show that the algorithm presented has higher spectrum efficiency compared with RBFN-based estimator using comb-type pilots. It is also found that this algorithm offers better performance over several linear-filtering-interpolation-based estimators. Secondly, the partition of received pilots is a problem when LS channel estimation is used in MIMO-OFDM systems. This dissertation proposes an LS-based universal orthogonal pilot design criterion. A zero-forcing pilot design method is then devised according to this criterion. Furthermore, an orthogonal pilot design method based on Hadamard matrix is put forward for maintaining the spatial diversity gain of pilot symbols. The orthogonal pilot design schemes and the channel estimation for pilots are given in detail according to diverse pilot types. Thirdly, the RBFN-based channel estimator for OFDM is introduced to MIMO-OFDM systems combined with the orthogonal pilot design method presented for MIMO-OFDM. At the same time, the specific designs for STBC-OFDM system with Alamouti Space-Time Block Codes are put forward. The simulation results show that this algorithm has similar advantages of performance and spectrum efficiency just as those of RBFN estimator in OFDM systems. Finally, this dissertation proposes a low complexity adaptive modulation scheme for VBLAST-OFDM system. An adaptive various step method is first employed to simplify a near-optimal adaptive algorithm used in OFDM with MQAM. It is generalized to VBLAST-OFDM after using SVD. The eigenvalues of channel matrix that represent the subchannel gains are employed in the optimal design of sub-channel's bit-loading scheme. The simulation results show that the presented scheme has a close performance to the sub-optimal algorithm but with more than 30% reduction in computation complexity.