In the era of the multimedia and high speed data transmission, one technique, multicarrier modulation (MCM), has been recognized to be the backbone of many broadband communication systems for both wire line and wireless environments. The success of MCM is due to the advances in digital signal processing and communication techniques. In this thesis, we address the channel estimation problem in the MCM systems. Our goal is to design a channel estimation scheme that is robust to environment variations. We achieve this goal by exploiting the unique feature of MCM and the correlation of the channel.; In this thesis, a channel estimation algorithm is proposed that uses the cyclic prefix that is specifically designed in MCM to reduce the inference between MCM blocks to estimate the channel responses. Our simulation shows that the algorithm can track the channel variation and hence improve the robustness of the system without additional training. Our performance analysis further proves the convergence of the algorithm.; Then a polynomial model based method is proposed to capture the correlation of the wireless fading multipath channel and hence reduce the noise in estimation. The method suppresses the leakage effect compared to the Fourier transformed based approach. It can adapt the estimator to the correlation of the channel so that we do not need to know the fading or the multipath characteristics to design the estimator.; The model based idea is further extended to solve the channel estimation problem in transmitter diversity systems. We investigate the identification and the training sequence design problems in the MCM systems with transmitter diversity and use the model based channel estimation approach in a robust progressive image transmission scheme.; Finally, we recognize that there remain many open areas that require more investigations and summarize some of the possible future research topics.
展开▼
