OFDM Channel estimation based on Gaussian Radial basis function interpolation

In this paper, we propose the new channel estimation algorithm for orthogonal frequency division multiplexing (OFDM) system in fast fading channels. It is two-step procedure. In the first step, we obtain an initial estimation of channel response at pilot subcarriers by least square method, then we employ the Gaussian radial basis function as a efficient nonlinear interpolator to estimate the channel response at the data symbol position. In the second step, we use the processing and filtering on the estimated channel taps based on magnitude or energy of channel coefficients. This step of algorithm is very effective in removing the LS estimation error.