DFT-based offset-QAM OFDM for optical communications

TLDR: It is shown that offset-QAM OFDM exhibits negligible penalty for all investigated spectra, in contrast to rectangular- function based Nyquist FDM (N-FDM) and sinc-function based conventional OFDM (C-OFDM), and by avoiding the GI, 112-Gb/s polarization multiplexed offset-4QAM OfDM can achieve 23% increase in net data rate over C-OF DM under the same transmission reach.

Abstract: We experimentally demonstrate and numerically investigate a discrete-Fourier-transform (DFT) based offset quadrature-amplitude-modulation (offset-QAM) orthogonal frequency division multiplexing (OFDM) system. We investigate the scheme using a set of square-root-raised-cosine functions and a set of super-Gaussian functions as signal spectra. It is shown that offset-QAM OFDM exhibits negligible penalty for all investigated spectra, in contrast to rectangular-function based Nyquist FDM (N-FDM) and sinc-function based conventional OFDM (C-OFDM). The required guard interval (GI) length for dispersion compensation in offset-QAM OFDM is analyzed and shown to scale with twice the subcarrier spacing rather than the full OFDM bandwidth. Experimental results show that 38-Gb/s offset-16QAM OFDM supports 600-km fiber transmission with negligible penalty in the absence of GI while a GI length of eight is required in C-OFDM. Further numerical simulations show that by avoiding the GI, 112-Gb/s polarization multiplexed offset-4QAM OFDM can achieve 23% increase in net data rate over C-OFDM under the same transmission reach. We also discuss the design of the pulse-shaping filter in the DFT-based implementation and show that when compared to N-FDM, the required memory length of the filter for pulse shaping can be reduced from 60 to 2 in offset-QAM OFDM regardless of the fiber length.