TCM-16QAM in nonlinear fiber transmissions

Abstract

We investigate the advantages and challenges associated with the implementation of trellis-coded modulation 16-ary quadrature amplitude modulation (TCM-16QAM) in optical fiber transmissions. Despite the potential benefits of enhanced sensitivity and spectral efficiency with TCM-QAM, fiber nonlinearity can compromise its performance at higher launch powers into optical fibers. This is mainly attributed to nonlinear noise and inter-symbol interference exacerbating burst errors post-Viterbi decoding. To address this challenge, we propose a low-complexity symbol-level block interleaving technique based on a simple transpose operation. This method disperses consecutive erroneous symbols into isolated ones, thereby reducing burst errors output from the Viterbi decoder. We first validate the proposed symbol-level method through simulations of polarization-division multiplexed (PDM) signals over 12 × 80-km standard single-mode fiber (SSMF). TCM-16QAM with interleaving depths of 2, 4, and 16 achieves a coding gain of 0.76–1.08 dB at a bit error rate (BER) of 1 × 10⁻³, while the launch power tolerance increases from 1.39 to 1.97 dB in the nonlinear transmission region compared to conventional 8 phase shift keying (8PSK) modulation with the same spectral efficiency. Furthermore, we perform experimental transmissions employing 32-GBaud PDM signals over 3 × 80-km SSMF. The experimental results reveal that in the nonlinear transmission region, TCM-16QAM with interleaving depths of 4 and 16 can achieve the launch power tolerance improvements of 0.48 dB and 0.67 dB, respectively.