minimum must be within the power spectrum’s envelop which yields [143] (for
Gaussian pulses) D>2.3Dt, where D is DGD and Dt is the pulsewidth.
In summary, DGD monitoring through XPM and optical spectral analysis will
enable closed-loop DGD compensation at ultrafast data rates. Future improve-
ments include reducing device length by using the high-nonlinearity glass fiber
and waveguides discussed.
In conclusion, all of these different approaches to OPM can be broadly divided
into phase-matched and nonphase-matched. All have advantages and disadvan-
tages—there is not necessarily a single best solution. They all rely on a nonlinear
PTF and so have the same basic principles and performance. The NOLM-based
methods offer some degree of flexibility in