Optimal OFDM Channel Estimation Under Different Channel Conditions

High-speed and high-quality services are demanded in modern mobile communication. Orthogonal Frequency Division Multiplexing (OFDM) has high level of spectral efficiency, combats multi-path fading effectively, and suits for wide-band mobile communication, and this makes it a promising technique in the next generation communication system. Channel estimation has attracted widespread attention, among which pilot symbol aided modulation (PSAM) is a practical channel estimation method in tracking the fast fading channel with low complexity and high accuracy.First, wireless channel is introduced with some primary characters, such as path loss, shadow fading, multi-path fading and Doppler frequency shift, etc. Then we propose the method of setting up channel model of Rayleigh fading channel and Rice fading channel. The Jakes model is used for the simulation of multipath time-variant channel.Then we give some general background information of OFDM algorithm, such as QPSK modulation, IFFT, cyclic prefix, D/A converter and so on. The principles of OFDM parameter selection are presented after analyzing its advantages and disadvantages. With the simulation results of OFDM signal in time-domain and frequency-domain, the performance and characteristics of OFDM are disclosed.We introduce some Pilot-Symbol-Assisted Modulation (PSAM) channel estimation techniques, like Least-squares (LS) and Minimum Mean Squared Error (MMSE) methods on pilot, and traditional interpolation algorithms, FFT transform domain, low-pass filter and DCT methods on other positions. On the basis of performance comparison of those methods under different channels, finally, we propose several optimal channel estimation methods applicable for different channel conditions.