Research On Enhanced Channel Estimation Technique For OFDM System

As one of the effective future mobile techniques, orthogonal frequency division multiplexing (OFDM) can not only effectively combat multi-path fading, but also strongly combat noises and inter symbol interfere (ISI).In this thesis, the principle of OFDM system is discussed based on the introduction of wireless fading channel. With focus on the key technique of channel estimation in OFDM systems, methods based on frequency-domain pilot and blind estimation are deeply researched. According to the current basis above, this thesis researches an enhanced algorithm of channel estimation in OFDM systems implemented in a low SNR transmission environment. In this algorithm, PN sequence as training sequence is inserted into OFDM symbols in time domain. Based on the property of pseudo-random (PN) sequence, the cross-correlation between the PN sequence and its corresponding received sequence is performed. Then the channel frequency response can be acquired through interative parameters detection. PN sequence has the same function as cyclic prefix (CP) because of its periodic appearance, and can also be used for synchronization. So this algorithm has high estimation accuracy and moderate complexity. Theoretical analysis and computer simulations show that, in multi-path Rayleigh fading channels, when SNR is lower, this algorithm can implement more accurate estimation than the algorithm based on frequency-domain comb pilot by 5dB.