WLC37-4: ICI Coefficient Estimation for OFDM Systems in Mobile Channels

Due to the superior spectral efficiency and robustness over the multipath channels, orthogonal frequency division multiplexing (OFDM) has served as one of the major modulation schemes for the modern communication systems. In the future, the wireless OFDM systems are expected to operate at high carrier-frequencies, high speed and high throughput mobile reception, where the fast time-varying fading channels are encountered. The channel variation destroys the orthogonality among the subcarriers and leads to the intercarrier interference (ICI). ICI poses a significant limitation to the wireless OFDM systems. In this paper, we develop a pilot-aided minimum mean square error (MMSE) ICI coefficient estimation algorithm for combating the rapid time-variant channels. Our MMSE ICI estimator utilizes the channel correlation properties. Since the channel statistics are usually unknown, we also investigate the mismatch between the underlying channel model and the actual channel statistics. Monte Carlo simulation results show that the achieved mean square ICI estimation error mostly depends on the number of pilots instead of such a mismatch. In addition, a MMSE equalizer incorporated with our ICI coefficient estimation scheme can greatly improve the symbol error rate in the rapid time-variant channels with multiple Doppler frequencies over the conventional one-tap equalizer.