A DFT-Based Channel Estimation Algorithm with Noise Elimination for Burst OFDM Systems

Accurate channel estimation can greatly improve the performance of burst OFDM system by compensating channel fading. Channel estimation based on discrete Fourier transform (DFT) is suitable for burst OFDM systems due to its low complexity and short convergence time. Traditional DFT-based channel estimation algorithm does not consider the noise influence within the length of cyclic prefix, which greatly decreases its performance. A DFT-based channel estimation algorithm with noise elimination (DCEA/NE) is proposed to estimate the instantaneous signal-to-noise ratio (SNR) using the training sequence in the burst OFDM system, and based on this, eliminate the noise influence. Simulation results show that the proposed algorithm outperforms the traditional DFT-based channel estimation algorithm.