The statistics of Intrachannel Four Wave Mixing IFWM

Outline Kerr nonlinearity induced nonlinear phase noise in coherent communication systems l Statistics of

Phase noise in coherent communication systems l Laser phase noise • Laser linewidth •

$Kerr Nonlinearity l induced intensity dependent refractive index l Self phase modulation induced Nonlinear$

Signal propagation in optical fibers Pulse trains l Nonlinear Schordinger Equation (NLSE) l l

What we know about IFWM l l IFWM technically information, but hard to fully

Real part negatively correlated Empirical distribution 1 pic of Liu corr… 1 pic of

Correlation of IFWM induced phase noise for QPSK systems

Variance goes with N^2 Fiber overall length L with N spans Opt. Amp.

Typical fiber types and dispersion maps (db/km) D(ps/nm-km) S(ps/nm^2 -km) SMF 0. 25 DCF

Multiplicative perturbation to NLSE l NRZ pulse with power 0 d. Bm, 45 km

Conclusions/Future Work l l Verify analytical predictions through simulation results Quantify the benefits of

Slides: 14

Download presentation

The statistics of Intra-channel Four Wave Mixing (IFWM) in Coherent Communication Systems Alan Pak Tao Lau Department of Electrical Engineering, Stanford University June 6, 2007

Outline Kerr nonlinearity induced nonlinear phase noise in coherent communication systems l Statistics of IFWM phase noise l Phase noise reduction through exploiting correlation of IFWM l Ongoing/Future work l

Phase noise in coherent communication systems l Laser phase noise • Laser linewidth • Carrier recovery mechanisms l Linear phase noise • ASE noise from inline amplifiers l Nonlinear phase noise • Phase fluctuations from randomness of data • Interaction of ASE noise and signal with Kerr nonlinearity l Shot noise / Thermal noise

$Kerr Nonlinearity l induced intensity dependent refractive index l Self phase modulation induced Nonlinear$

Kerr Nonlinearity l induced intensity dependent refractive index l Self phase modulation induced Nonlinear Phase Shift

Signal propagation in optical fibers Pulse trains l Nonlinear Schordinger Equation (NLSE) l l Pertubation l l l SPM on IXPM on IFWM on Linear solution to NLSE : : :

What we know about IFWM l l IFWM technically information, but hard to fully exploit (requires complexity exp(K) for optimal detection) IFWM phase noise are correlated Wei and Liu, Optics Lett. , Vol. 28, Issue 23, pp. 2300 -2302 Ho, PTL vol. 17, no. 4, Apr. 2005, pp. 789 -791 l Basically, know nothing about IFWM

Real part negatively correlated Empirical distribution 1 pic of Liu corr… 1 pic of Ho empirical pdf

Correlation of IFWM induced phase noise for QPSK systems

Variance goes with N^2 Fiber overall length L with N spans Opt. Amp.

pdf

Pairwise independent

Typical fiber types and dispersion maps (db/km) D(ps/nm-km) S(ps/nm^2 -km) SMF 0. 25 DCF 0. 6 ~17 0. 25 Typical dispersion maps: l SMF/DCF l NZDSF-/SMF l SMF/IDF 1. 3 ~-0. 43 NZDSF 0. 25 IDF ~0. 06 (/W-km) ~-40 ~0. 05 ~2 ~-0. 13 ~3

Multiplicative perturbation to NLSE l NRZ pulse with power 0 d. Bm, 45 km of NZDSF D=-2 ps/nm-km,

Conclusions/Future Work l l Verify analytical predictions through simulation results Quantify the benefits of sequential detection and pulse shape design Acknowledgements l Sahand Rabbani