Clipping noise suppression in OFDMA using frame based alternating projection

In OFDMA systems, receiver-oriented methods present new challenges in suppressing clipping noise to compensate for power amplifier nonlinearity because not all modulation information about other users is available to any individual user. Recently a scheme based on band limited signal recovery has been proposed to address this problem but its performance in high SNR regime is limited. In this paper, we cast the problem as one of the clipping noise suppression estimation using the framework of frame theory together with alternating projection over convex sets to improve the performance. Extensive simulations show the proposed scheme outperforms existing schemes without increasing complexity, and in particular, system performance in high SNR regime is significantly improved.