Channel estimates based on frequency domain interpolation for OFDM systems

Coherent orthogonal frequency division multiplexing (OFDM) systems need to accurately estimate the channels. This paper presents a channel estimator which suppresses the noise during frequency interpolation. Discrete Fourier transforms (DFT) are used to estimate the channel impulse response using frequency domain plots. Weighting of the channel impulse response according to its envelope then reduces the channel estimation errors caused by noise. The weighting factor can be simplified to reduce hardware complexity. Simulation results show that the method outperforms the traditional DFT-based method and linear interpolation at low signal to noise ratios (SNR), but that the advantages of the method decrease as the SNR increases. The method is quite simple, so it can be easily realized in hardware.