Research On Channel Estimation For OFDM

Channel estimation is the critical part of OFDM system, its good performance is the basic requirement of system performance. The future mobile communication will have become deeply integrated into people’s lives, and the communications environment will become more complicated, so it is practical and academic to study a high-performance and strong robustness channel estimation algorithm.In OFDM based systems, the original data is transformed to several parallel data streams, then modulated onto the orthogonal sub-carriers. With a degradation rate of one data stream, the width of one symbol has reduced and making it robust and more resistant to interference, the orthogonal sub-carriers, on the other hand, improves the bandwidth efficiency. In this paper, we mainly study the channel estimation based on pilot tones. Firstly, we analyze the principals of OFDM technologies and propose the baseband system model. Then, we study the large-scale fading and small-scale fading, and present a simulation on common channel model, analyze the statistics characteristics and affecting factors of wireless channel. Secondly, we study the principals of channel estimation based on pilot tones, especially research on pilot structures, channel estimation criteria. Finally, we mainly analyze transform domain based channel estimation, and propose a novel channel estimation algorithm based on QR matrix decomposition to reduce estimation error. The simulation results in the Doppler channel show that the bit-error-rate(BER) of our estimator is2-3dB better than the conventional DFT-based channel estimator and LS-linear estimation.Channel estimation performance has a direct influence on OFDM system, which has been considered to be one of the key technologies for next generation mobile communications. In this paper, a novel domain transition algorithm has been proposed, to improve performance of OFDM.