| Recent researches show that, using multiple antennas on both transmitter and receiver sides of wireless communication systems, MIMO techniques can provide space multiplex and space diversity gains. Space multiplex gain may effectively increase the channel capacity and spectral efficiency, and space diversity gain may help wireless communication systems to combat multi-path scattering effect of the wireless channel. On the other hand, OFDM which enssentially is one kind of multi-carrier modulation techniques can extremely increase the spectral efficiency and combat inter-symbol interference (ISI) which is generated by frequency-selective wireless channel. Therefore, MIMO technique being combined with OFDM wireless communication systems into MIMO-OFDM system holds many advantages, such as additional spatial diversity, high spectral efficiency, combating ISI, simple construction and so on. Therefore, MIMO-OFDM is well known to be one of the main physical layer techniques in the future wireless communication standards.Like the single antenna OFDM systems, however, MIMO-OFDM wireless systems are also very sensitive to the carrier frequency offset (CFO) and phase noise (PN). CFO and PN could destroy the orthogonality between sub-carriers of MIMO-OFDM systems, thus bring severe inter-carrier interference (ICI) and significantly degrade bit error rate (BER) performance of the MIMO-OFDM systems. In addition, channel estimation is mandatory in the receiver of MIMO-OFDM systems to achieve the high data rate. Therefore, channel estimation and frequency synchronization for MIMO-OFDM systems are hot research topics in the field of wireless communications.The dissertation discusses several channel estimation and frequency synchronization methods for MIMO-OFDM systems, and aim to construct the research framework of the channel estimation and frequency synchronization for MIMO-OFDM systems. By utilizing stochastic signal processing and adaptive signal processing approaches, author has done some researches on time-varying channel estimation in MIMO and MIMO-OFDM systems, CFO estimation in OFDM systems, joint channel and CFO estimation in MIMO-OFDM systems, and joint channel, CFO and PN estimation in MIMO-OFDM systems, respectively. The main contributions of this dissertation include:(1) Particle filtering based semiblind channel estimation for MIMO-OFDM systems. Based on the particle filtering theory, a semiblind channel estimation method is proposed for MIMO-OFDM systems. Multi-object optimization is used to improve the sequential Monte Carlo (SMC) particle filter (PF). Then, a time-varying channel estimation method for MIMO-OFDM systems is presented by using improved SMC PF. This method has good channel estimation performance, thus improves the symbol detection performance of MIMO-OFDM systems. Simulation results show that, this method is robust to non-Gaussian noise, and performs well for both slow and fast time-vary ing channels.(2) Time-varying channel Bayes estimation based on Wishart random matrix for MIMO systems. Based on the random matrix theories, Gibbs sampling algorithm is used to jointly estimate the channel and covariance matrix of the noise, whose conjugate prior distributions are selected by complex Gaussian and inverse-complex Wishart distributions, respectively. Then, time-varying channel can be tracking using SMC PF which is improved by modifying the candidate distribution. Simulation results show that, this method could effectively improve the channel estimation and symbol detection performance of MIMO systems in the color noises.(3) H∞filtering based channel estimation for STBC-OFDM systems. Based on the H∞filtering theory, a time-varying channel estimation algorithm is proposed for STBC-OFDM systems. By utilizing the orthogonal code construction of space-time coding, this algorithm holds low computational complexity. Moreover, channel and noise statistics are not required in the design of this algorithm, which thus is robust to the unknown disturbance such as noise and model errors. Simulation results show the effectiveness of the proposed algorithm.(4) H∞particle filtering based time-varying channel and CFO estimation for MIMO-OFDM systems. An H∞particle filtering algorithm is proposed by combining Rao-Blackwellized (R-B) particle filter with H∞filtering. Using this H∞particle filtering algorithm, time-varying channel and CFO could be jointly estimated in MIMO-OFDM systems. In the proposed algorithm, H^ filters are used to estimate the channel parameters corresponding to each CFO samples. And CFO could be estimated by using SMC algorithm. Simulation results show that, the proposed algorithm is robust to both non-Gaussian noise and model errors, hence improves the detection performance of MIMO-OFDM systems.(5) Extended H∞filtering based CFO estimation for OFDM systems. Based on the H∞filtering theory, a CFO estimation algorithm is proposed for OFDM systems. The design criterion of this algorithm is to minimize the worst effect of noise and model errors on CFO estimation errors. Compared with EKF, the proposed algorithm does not require the statistics of unknown noise, and is consequently robust to the model error of ambient noise. In addition, its computational complexity is also not higher than EKF. Simulation results show that, the proposed algorithm could combat the ICI resulted from CFO, thus improve OFDM system performance. (6) H∞filtering based CFO estimation for MIMO-OFDM systems. Based on the H∞, filtering theory, a CFO estimation algorithm is proposed for MIMO-OFDM systems. This algorithm uses unscented transformation to convert nonlinear CFO estimation to a linear H∞filtering problem, thus achieves iterative estimation of channel and CFO without statistical knowledge of the noise. Simulation results show that, this algorithm is robust to non-Gaussian noise and model error, thus improves the performance of reveiver.(7) Joint estimation of channel, CFO and PN in MIMO-OFDM systems. Based on the MAP estimation criterion, a joint channel, CFO and PN estimation algorithm is proposed for MIMO-OFDM systems. This algorithm could achieve accurate channel, CFO and PN estimation, only using one OFDM symbol as preamble training sequence. Simulation results show that, the proposed algorithm can effectively improve the channel estimation performance in the presence of CFO and PN, combat the ICI resulted from CFO and PN, and improve the performance of MIMO-OFDM systems.
|
PDF Full Text Request