Iterative Partial-Interference-Cancellation-based Detector for OFDM Systems over Doubly-Selective Rayleigh Fading Channels

The time-variation of channels during one OFDM symbol destroys the orthogonality among subcarriers and results in intercarrier interference (ICI), which may detail substantial error performance. Assumed channels vary in linear fashion during one OFDM symbol, we propose an iterative partial-ICI-cancellation (PIC) based detector for OFDM systems with linear modulation over doubly-selective Rayleigh fading channels. The tentatively detected information of channel and data are fed back to regenerate the ICI and then partially subtracted from the received signals. The optimal partial ICI cancellation factor (PICF) is determined by minimizing the mean square error (MSE) of data estimation. The proposed detector iteratively renews the channel estimation and data detection after PIC, and results in performance enhancement. Based on Jake's model for the Doppler effects, the bit error ratio (BER) performance of the proposed detector is illustrated for OFDM systems with phase-shift keying (PSK) and quadrature-amplitude modulation (QAM) in Rayleigh fading channels at several speeds.