A robust pilot-symbol-aided channel estimator for OFDM wireless communications

This paper presents a pilot-symbol-aided channel estimator for orthogonal frequency-division multiplex (OFDM) wireless communication systems. The proposed scheme is highly robust to time variation of wireless channels. In addition, it is equivalent to the conventional two-dimensional (2-D) minimum mean square error (MMSE) channel estimator, but with greatly reduced computational complexity. The reduction in complexity is achieved by employing the 2-D inverse fast Fourier transform (IFFT), 2-D FFT, and a 2-D weighting function instead of a 2-D filter. The weighting function is derived based on the mean square error (MSE) criterion and is simple to implement. For cases where channel statistics are not available, a robust estimator based on a simple 2-D windowing function is proposed. Furthermore, an enhanced channel estimator that can further improve the performance of the robust estimator is proposed. Simulation results demonstrate that the proposed robust and enhanced estimators are highly effective for realistic situations without the need for channel statistics.