Channel Estimation And Interference Cancellation Techniques In OFDM System

OFDM system is widely applied in wireless broadband communications as it has high frequency efficiency and good performance over frequency-selective channels.The increasing of symbol duration makes OFDM more robust against multipath transmission, whereas, it makes OFDM more sensitive to the time varying of wireless mobile channels. Doppler frequency expansion due to the channel variation destroys the orthogonality among different subcarriers, causes power leakage and introduces intercarrier interference (ICI). The ICI degrades the performance of OFDM systems and makes channel estimation more challenging as well.The objective of this thesis is to propose new solutions and algorithms related on OFDM channel estimation and interference cancellation techniques over time-varying channels based on the literatures, and analyze OFDM system based on the framework of filterbank theory.The contributions of this thesis are as follows:(1) Propose two algorithms for channel and ICI coefficients estimation in OFDM system over approximately linear channel (channel characteristics in time-domain vary in an approximately linear way during one OFDM symbol time), one is in frequency-domain, the other is in time-domain. The parameters need to estimate in time-domain method are less than those in frequency-domain method. The simulation results show that the time-domain method is more robust when Doppler frequency is high.(2) Study the ICI in detail. Derive the upper and lower bounds of ICI power on Jakes channel model, analyze the effects of the ICI power distribution on the performance of OFDM system under different channel conditions with MATLAB simulations.(3) With the ICI analysis, Modify the MMSE criterion over time-varying channel and propose a novel parallel iterative equalization method to mitigate the ICI. Simulation results demonstrate that the proposed method outperforms the traditional parallel iterative cancellation method in BER performance, but has relatively slower convergence speed.(4) Propose a direct ICI mitigation solution based on the LMMSE criterion which is both applicable over the channels varying in slow speed and in high speed. We derive the general formulation on time-varying channel, and the special case on Jakes channel as well. Simulation results illustrate that this solution has robustness against the variation speed. But if there is a mismatch of channel model, the performance will degrade.(5) Propose a blind and equalization method based on a nonmaximally decimated filterbank precoder and give the expressions of the optimal equalization solution and the SNR on the output end of the channel. The great strength of this method is that it can estimate unknown FIR channels without limitations of the location of zeros.(6) Analyze the OFDM system under the framework of filterbank theory. Present a general system model for OFDM in both single user and multiuser scenarios based on the filterbank-constructed transmultipexer structures.