Analysis and Experimental Demonstration of Novel 8Polsk-Qpsk Modulation at 5 Bits/symbol for Passive Mitigation of Nonlinear Impairments.
Mathieu Chagnon,Mohamed Osman,Qunbi Zhuge,Xian Xu,David V. Plant
DOI: https://doi.org/10.1364/oe.21.030204
IF: 3.8
2013-01-01
Optics Express
Abstract:We propose a new modulation format providing 5 bits of information per recovered symbol while maintaining constant the total optical power. The proposed format applies a simple power constrain to the DP-8QAM format. This modulation format provides a passive way to mitigate nonlinear impairments due to Kerr effects occurring during propagation, and most specifically in the first 40 km. This report presents, to the authors' knowledge, a new transmission format using solely phase and polarization as modulation methods. The performance of this format, named 8PolSK-QPSK, is experimentally compared with that of the DP-8QAM format as both require equal transmitter complexity and implementation penalty, at the expense of a 20% increase in signaling baud rate. The greater nonlinear tolerance of this format is experimentally demonstrated. Moreover, thorough analysis of the Manakov-PDM propagation equation applied to both formats provides analytic explanation of the 8PolSK-QPSK's improved performance. The constant power property of the symbol set of the proposed format mitigates self- and cross-phase modulation (SPM, XPM) nonlinear effects and is experimentally validated over a long-haul transmission system in a WDM scenario. A total throughput of 7 × 129 Gbps is maintained for the transmission format comparison. Simulation of the same transmission system allows separate analysis of the strength of SPM, XPM and Cross-Polarization Modulation (XPolM) nonlinear effects and demonstrate reduced nonlinear impairments for the proposed format in the first span. We show an experimental reduction of the required OSNR for a BER threshold of 1.4 × 10(-2) of 0.5 dB for 8PolSK-QPSK compared to DP-8QAM in back-to-back. After 1920 km of SMF fiber, we demonstrate a required OSNR (ROSNR) diminution for increasing launch power, allowing a ROSNR relief of 0.95 dB at optimal launch power of -1 dBm for the proposed format. Using the same threshold, we show an increased reach by more than 34%, or 975 km, at optimal launch power. We also demonstrate that the relative reach increase for 8PolSK-QPSK compared to DP-8QAM monotonically increases with increasing BER threshold and that the BER growth with distance, after the first span, is equal for both formats.
What problem does this paper attempt to address?