Channel Estimation In Wireless OFDM Communication Systems

Due to its features of multipath immunity and bandwidth efficiency, OFDM (Orthogonal Frequency Division Multiplexing) has been favored for both Europe DAB (Digital Audio Broadcasting) and DVB (Digital Video Broadcasting), and it has become a prime candidate for B3G (Beyond 3rd Generation) communications. The requirements for B3G systems such as high data rate, high quality and high mobility make channel estimation and signal detection techniques more strict; hence we focus on channel estimation and joint channel estimation and symbol detection.We investigate the channel estimation algorithms based on tracking multipath time delay, a low complexity multipath time delay estimation method is proposed. The theoretical analysis and simulation results show that the dimension of estimated parameters can be effectively decrease by tracking multipath time delay, consequently, the channel estimation performance is improved. Simulation results for MIMO-OFDM (Multiple Input Multiple Output OFDM) systems demonstrate that the proposed method can track the multipath time delay information more quickly.The usage of pilot symbols increases the system overhead and reduces the bandwidth efficiency, whereas, semi-blind channel estimation and iterative estimation and detection are effective techniques to improve bandwidth efficiency, so the semi-blind channel estimation using superimposed training in time domain is investigated. We derive the MSE (Mean Square Error) performance of channel estimation employing a first-order statistic. Based on these results, we analyze the different optimization criterions for superimposed pilot and design the corresponding training sequences. Furthermore, the iterative joint channel estimation and signal detection is applied to improve the system performance.The design of iterative receivers with adaptive training scheme for SFBC-OFDM (Space Frequency Block Coded OFDM) systems is considered. We propose a simple criterion for error propagation detection, based on which, the training interval is adjusted adaptively. Simulation results show that compared with fixed training interval schemes, the improved algorithm can achieve better BER (Bit Error Rate) performance while keeping lower system overhead.After nonideal frequency offset compensation and symbol synchronization, residual frequency and time offset may influence the channel estimation performance. We derive the MSE performance of the pilot aided channel estimation under nonideal frequency and time synchronization. The theoretical conclusions are verified by...