Impact of Chromatic Dispersion Compensation in Single Carrier Two-Dimensional Stokes Vector Direct Detection System

We experimentally demonstrate, aided with mathematical analysis, two methods for chromatic dispersion (CD) compensation for two-dimensional Stokes vector direct detection (2D-SV-DD) system. Results show that the better CD compensation method is dependent on transmission distance and operating symbol rate relative to the digital-to-analog converter (DAC) sampling rate. When 84 GBd 16 QAM signal is generated using an 84 GSps DAC, the bit error rate (BER) after 640 km propagation is $\sim1.67\times 10^{-2}$ and $\sim2\times 10^{-2}$ using CD post- and precompensation, respectively. When 64 GBd 16 QAM signal is generated at the same DAC sampling rate, the BER at 960 km is $\sim1.61\times 10^{-2}$ and $\sim9.8\times 10^{-3}$ using CD post- and precompensation, respectively. In addition, our findings reveal that CD precompensation allows for transmission distances greater than 1000 km at which CD postcompensation is no longer viable due to the required training. Finally, enabled by CD precompensation, we report the highest capacity-times-distance products of 591 360 (Gb/s) $\cdot$ km and 614 400 (Gb/s) $\cdot$ km in 2D-SV-DD system using QPSK modulation format and symbol rates of 84 and 64 GBd, respectively. Also, we can achieve 300G transmission for distances below 320 km using 84 GBd 16 QAM signal with BER below hard decision forward error correcting threshold of $\sim3.8\times 10^{-3}$ .