Semi-Blind Channel Estimation and Multi-User Data Detection for Downlink MC-CDMA

This paper proposes a unique low complexity semi-blind channel estimation and data detection receiver for downlink MC-CDMA systems. In this iterative scheme, the CSI is predicted through short range wireless channel prediction algorithm initially. With the predicted CSI, the novel data detection based on PIC is employed to achieve the data decision results. The refined CSI is acquired from the decision feedback simultaneously. This process repeats multiple times until the CSI converges. With the use of QMMSE/MMSE-PMRC PIC based on CQS instead of Walsh codes, the performance of channel estimation and Multi-User Detection is improved and the system complexity is reduced at the same time. The simulation results indicate that our proposed system has a good BER performance and channel tracking ability in different channels and system loading conditions.