Efficient Intercarrier Interference Mitigation for Pilot-Aided Channel Estimation in OFDM Mobile Systems

Motivated by the possibility of decreasing the intersymbol interference (ISI) which is due to large delays of a multipath mobile radio channel, orthogonal frequency division multiplexing (OFDM) became very popular. However, the timevariance of the mobile radio channel induces intercarrier interference (ICI) yielding substantial channel estimation errors and thereby tremendous transmission impairments. Contrary to previous algorithms which resort to a linearization of the time-variant channel, we combat the ICI using eigenspaces of time-domain covariance matrices defined by the autocorrelation function of the Doppler spread. We perform a basis expansion using Slepian sequences and determine the basis coefficients of the time-variant channel by channel estimates from previous OFDM symbols. Once we know these basis coefficients, we obtain the necessary time-variant channel estimation by the Slepian sequences. These time-variant channel estimates allow a symbol detection in frequency domain which eliminates the ICI almost completely. Simulation results investigate both the signal to interference ratio (SIR) and the bit error ratio (BER) of our new ICI mitigation methods and reveal the superiority compared to previous algorithms for ICI reduction.