The document analyzes the impact of channel add/drop on nonlinear performance in uncompensated 120 Gb/s PM-QPSK links. It shows up to 1.25 dB of nonlinear tolerance benefit when all neighboring channels are added and dropped every 2 spans compared to no add/drop. However, pre-dispersion of added channels reduces this benefit. Further study of interactions with optical filtering and different fiber types is needed.

1. Steven Searcy and Sorin Tibuleac
ADVA Optical Networking, Norcross, GA, USA
14 October 2014
Nonlinear Impact of Diverse
Optical Routing in Uncompensated
120 Gb/s PM-QPSK Links

2. IEEE Photonics Conference 2 2014, paper TuF1.4
• Assessing transmission performance for DWDM systems
• Common assumption in analyzing system performance (esp. nonlinear
effects) is that all channels co-propagate over entire link
• Many deployed point-to-point systems match this configuration
Introduction & Motivation
Purely
Point-to-Point
System
(No Add/Drop)

3. IEEE Photonics Conference 3 2014, paper TuF1.4
• Assessing transmission performance for DWDM systems
• Common assumption in analyzing system performance (esp. nonlinear
effects) is that all channels co-propagate over entire link
• Many deployed point-to-point systems match this configuration
• However, many deployed systems have a substantial degree of add/drop
traffic and may behave differently  investigate this scenario
Introduction & Motivation
Network with
Add/Drop
traffic at
ROADM nodes
along path

4. IEEE Photonics Conference 4 2014, paper TuF1.4
• Changes in add/drop conditions have been shown to have
substantial impact on nonlinear tolerance in CD-compensated
10G systems; effect is strongly dependent on residual CD
Dispersion compensated systems
Figure from [3] T. Zami, et al., Proc. ECOC 2009, paper 1.5.2

5. IEEE Photonics Conference 5 2014, paper TuF1.4
• Nonlinear performance in uncompensated system often analyzed
based on assumption of Gaussian nonlinear noise [1]
• These underlying conditions are not always met with frequent add/drop
• Add channels may have zero initial CD if originating at Add node, or
some positive CD accumulated on a different path in mesh network
Dispersion uncompensated systems
Figure from [5] F. Vacondio, et
al., Opt. Express 20(2), 2012.

6. IEEE Photonics Conference 6 2014, paper TuF1.4
Experimental configuration
• Commercial 120 Gb/s PM-QPSK transceiver with real-time DSP
• Recirculating Loop with all-EDFA amplification, four spans
TrueWave-RS fiber, WSS every two spans for optional add/drop

7. IEEE Photonics Conference 7 2014, paper TuF1.4
Test conditions
• 19x100G channels on 50 GHz grid (test channel + 18 neighbors)
• 50GHz channel slots adjacent to test channel are left empty, to eliminate
filtering on test channel in all cases  optical filtering may produce
complex interactions with nonlinear effects [4] (left for further study)
• Four different test cases:
• A: No Add/Drop (all channels co-propagate)
• B: Add/Drop all neighbors every 4 spans
• C: Add/Drop all neighbors every 2 spans
• C-PD: same as C, with Add channels pre-dispersed
• Initially passed through 1 span of ULAF (~2000 ps/nm pre-dispersion)

8. IEEE Photonics Conference 8 2014, paper TuF1.4
Nonlinear OSNR Penalty results
• Baseline case—no add/drop, all channels co-propagate

9. IEEE Photonics Conference 9 2014, paper TuF1.4
Nonlinear OSNR Penalty results
• ~0.75-1.0 dB nonlinear benefit (@1 dB OSNR penalty)
from Add/Drop of all neighbors every 4 spans

10. IEEE Photonics Conference 10 2014, paper TuF1.4
Nonlinear OSNR Penalty results
• ~1.0-1.25 dB benefit from Add/Drop of all neighbors every 2 spans
 More frequent add/drop only produces incremental improvement (0.25dB)

11. IEEE Photonics Conference 11 2014, paper TuF1.4
Nonlinear OSNR Penalty results
• Pre-dispersion on add signals greatly reduces nonlinear benefit
Higher input CD degrades NL tolerance, esp. on NZ-DSF [1,5,6]

12. IEEE Photonics Conference 12 2014, paper TuF1.4
• Nonlinear benefit from frequent add/drop  up to 1.25 dB shown
for add/drop of all neighbors every 2 spans
• Much of the add/drop nonlinear benefit is due to reduced PAPR
when add signals have zero initial CD
• Topics for further study
• Behavior with combined nonlinearity and optical filtering
• Filtering penalty will at least partially offset nonlinear benefit?
• Behavior with other fiber types besides NZ-DSF
• Less impact from pre-dispersion over higher-dispersion fiber? (e.g. SSMF)
• Other modulation formats
• Mixed QPSK-OOK networks?
• Next-gen 16QAM/8QAM?
Conclusions

13. SSearcy@advaoptical.com
Thank You
Questions?

14. IEEE Photonics Conference 14 2014, paper TuF1.4
• [1] P. Poggiolini, G. Bosco, A. Carena, V. Curri, Y. Jiang, and F. Forghieri, “The GN-Model of Fiber Non-
Linear Propagation and its Applications,” J. Lightwave Technol. 32(4), pp. 694–721, February 2014.
• [2] T. Zami, A. Morea, and N. Brogard, “Impact of routing on the transmission performance in a
partially transparent optical network,” Proc. OFC/NFOEC, paper JThA50 (2008).
• [3] T. Zami, P. Henri, L. Lorcy, and C. Simonneau, “Impact of the optical routing on the transmission
in transparent networks,” Proc. ECOC, paper 1.5.2 (2009).
• [4] Y. Ye, G. Goeger, E. Zhou, S. Zhang, and X. Xu, “Interplay of Filtering and Nonlinear Transmission
in Coherent Uncompensated DWDM System,” Proc. OFC/NFOEC, paper OM3B.4 (2013).
• [5] F. Vacondio, O. Rival, C. Simonneau, E. Grellier, A. Bononi, L. Lorcy, J.-C. Antona, and S. Bigo,
“On nonlinear distortions of highly dispersive optical coherent sytsems,” Opt. Express 20(2), p. 1022-
1032, Jan. 2012.
• [6] X. Liu and S. Chandrasekhar, “Experimental Study of the Impact of Dispersion on PDM-QPSK and
PDM-16QAM Performance in Inhomogeneous Fiber Transmission,” Proc. ECOC, paper P.4.17 (2013).
• [7] S. Searcy and S. Tibuleac, “Impact of Channel Add/Drop on Nonlinear Performance in
Uncompensated 100G Coherent Systems,” submitted to OFC 2015.
References