Channel Estimation For Pilot Assisted OFDM System

OFDM(Orthogonal Frequency Division Multiplex)is a key technology of 3G. Because of its advantages such as high data rate, anti- multi path attenuation and frequency-selective attenuation and easy combination with other key technologies like MIMO, it has gained importance nowadays.Channel estimation techniques play an important role in the wireless mobile OFDM system, for its performance may affect that of the whole system. This thesis mainly focuses on the algorithm of channel estimation based on pilots.First of all, analyze the transmission characteristic of wireless channel, describe in detail several types of common attenuation and give the transmission model of wireless channel. Elaborate the principle and realization of OFDM. Summarize briefly the key techniques for OFDM.On that basis, research mainly the algorithm of channel estimation based on pilots. Generalize estimation method including LS, MMSE, LMMSE, and SVD. After that, introduce several interpolation filter including linear interpolation, cubic spline interpolation, Gaussian interpolation and low-pass filtering. To eliminate the ICI induced by time-varying channel, summarize some effective methods such as ICI self-cancellation, windowing, difference, piece-wise model interpolation, and compare their performance by simulation. Raise a Second-order Polynomial model to approximate the channel time-domain response in three consecutive symbols. Simulation results show that this method reduces the impact of ICI and enhances estimation performance greatly.