Channel Estimation And Equalization In OFDM-CDMA Systems

The aim of this paper is to investigate the problem of pilot-assisted channel estimation in downlink MC-CDMA. Channel estimation is divided into LS estimation at pilot tones and the interpolation of the preliminary estimate. This paper introduces a bispectrum-based channel estimation algorithm to improve the accuracy of the channel estimation at pilots and proposes an equalization method with the view of the reduction of the interpolation error.To overcome the affection of ICI and AWGN on the LS estimate of the transfer function at pilot tones, the proposed bispectrum-based channel estimation algorithm employ the suppression of ICI and AWGN components in the received pilot signals by bispectrum and then the accuracy of the estimation is improved. With the channel estimation at pilot tones, an efficient interpolation is necessary. An improved interpolation method based on transform-domain filtering is proposed, which employs DFT/IDFT and filtering in a transform domain so that the effect of noise on the equalization is significantly reduced.