Prediction of channel information in multi-user OFDM systems

Abstract—Channel information is indispensable to employ advanced channel aware technologies such as packet scheduling and adaptive modulation and coding (AMC). In this paper, we first investigate the delay effect of instantaneous signal to interference and noise power ratio (SINR) on the spectral efficiency of multi-user orthogonal frequency division multiplexing (OFDM) systems that employ packet-based channel aware technologies. To alleviate the performance degradation due to delayed channel information, we consider the use of predicted channel gain for link adaptation with packet scheduling. It is shown that the use of predicted channel gain can significantly enhance the spectral efficiency particularly in high mobility environments. For practical realization of an optimum predictor, we propose a grouped minimum mean square error (MMSE) prediction scheme, which can substantially reduce the implementation complexity without noticeable performance degradation. Finally, the proposed scheme is verified by computer simulation. Index terms: channel state information, packet scheduling, MMSE prediction, OFDM I