Clipping noise cancellation in uplink MC-CDMA system using signal reconstruction from non-uniform samples

In this paper, an iterative signal reconstruction method is extended to cancel multi-user clipping noise in a uplink MC-CDMA systems. Clipping is the simplest method to overcome high peak-to-average power ratio of multi-carrier signals but it makes the multi-carrier signals distorted. Reconstruction methods use non-distorted samples to reconstruct distorted samples in the receiver but multi-user interference causes the methods do not work properly because all the received samples are distorted by clipping and interference and so there is no undistorted samples to be used in recovering clipped samples. On the other hand, multi-user interference cancellation methods do not work properly because clipping in the transmitter is a non-linear process and therefore this process increases the amount of multiple access interference (MAI). In this paper, we propose a joint MAI and clipping noise cancellation method. In the proposed method in each iteration a MC-CDMA detector is used for signal detection and then using non-uniform sampling (that forces known error positions due to clipping to be zero) and filtering process an error estimation is calculated and then canceled. Simulation results show that the proposed method improves the performance of clipped MC-CDMA systems in uplink.