On The Synchronization And Channel Estimation For TDD Mode Based CDMA Systems

With nearly ten-year development of CDMA mobile communication, the FDD mode has attracted a lot of attention. However, FDD mode can be only used in symmetric frequency band in the mobile communications, thus it limits the utilization of available frequency bands. On the other hand, TDD CDMA has many technology advantages such as high spectrum effectivenss, asymmetrical uplink and downlink traffic supporting. This thesis is mainly paid to TDD CDMA vector signal generator and analyzer software with the main research work focused on the design and implementation of synchronization and channel estimation modules based on the MS Windows software platform, as well as the related theoretical work.This thesis briefly introduces the TDD CDMA communication theory at first, including CDMA spread spectrum communication principle, and the similarities and differences of the FDD mode and the TDD mode. By taking the LCR TDD as an example, the frame structure of TDD CDMA system and the process of downlink synchronization after UE powering on are studied. Besides, an improved synchronization algrithm under the large frequency offset, which can reach chip-level accuracy and provide the good guarantees to the coming procedure such as frequency offset estimation, is also proposed. In addition, to satisfy the demand of signal test of TDD CDMA signal analyzer, the synchronization module includes not only the pilot-based synchronization but also the midamble-based synchronization, which can be well used in mobile Ad hoc network based 3G technology for the practical applications.TDD CDMA system adopts joint detection, which significantly improves its capacity. But the performance of joint detection algorithm lies in the accurate channel estimation. In the fourth chapter, this thesis firstly investigates firstly the wireless channel model, where a midamble sequence in each burst is adopted for channel estimation. Then we compare low cost channel estimation with correlation algorithm. Because of interference causesd by multi-path and noise, the channel estimation is not accurate, which causes the inaccuracy of the joint detection. An improved algorithm is proposed, which sets two gates for the channel impulse responses, and thus can provide more accurate values for joint detection receiver.In the fifth chapter, according to 3GPP protocol, this thesis compares the system performance of different channel parameters and of different channel estimation methods. Finally, some useful conclusions are obtained and the advantages of the proposed algorithm are verified, together with problem analysis and experience summary on the process of thesis. Also some future directions are given.