Abstract
In this paper, a new decision-directed signal reconstruction (DDSR) algorithm is proposed for clipping noise mitigation in optical orthogonal frequency division multiplexing (OFDM). OFDM is increasingly being considered as the modulation method for the next-generation intensity modulation/direct detection (IM/DD) based optical transmission systems. When optical OFDM is used, clipping noise remains one of the main challenges. In previous studies, a time-domain based DDSR algorithm has been used to effectively mitigate the clipping noise. In this algorithm, based on the preliminary data detection decisions, an unclipped version of the OFDM signal is reconstructed non-linearly which can be decoded with better performance compared with the conventional OFDM signal detection method. However, due to the influences of the unavoidable noise at the receiver, many clipped signal samples can not be correctly processed which significantly limits the advantages of using this algorithm. In this paper, a new linear signal reconstruction method is proposed. Using this linear approach, all clipped signal samples can be successfully updated in the signal reconstruction process which consequently mitigates the clipping noise more effectively compared with the previous non-linear methods. Simulation results are presented to show that, when this new form of DDSR method is used, the bit error rate (BER) performance of the system is significantly improved compared with both the conventional receiver and the previous non-linear DDSR receiver.