29.03.2016 Views

Nonlinear Fiber Optics - 4 ed. Agrawal

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

10.4. Parametric Amplification 395<br />

where κ = Δk+γ(P 1 +P 2 ) and P 1 and P 2 are the input pump powers, assum<strong>ed</strong> to remain<br />

undeplet<strong>ed</strong>. The amplification factor is relat<strong>ed</strong> to g as indicat<strong>ed</strong> in Eq. (10.4.7) and is<br />

given by<br />

G s = P 3 (L)/P 3 (0)=1 +(2γ/g) 2 P 1 P 2 sinh 2 (gL). (10.4.15)<br />

Similar to the single-pump case, the gain spectrum is affect<strong>ed</strong> by the frequency<br />

dependence of the linear phase mismatch Δk. If we introduce ω c =(ω 1 + ω 2 )/2asthe<br />

center frequency of the two pumps with ω d =(ω 1 − ω 2 )/2, and expand Δk around ω c ,<br />

we obtain [89]:<br />

Δk(ω 3 )=2<br />

∞<br />

∑<br />

m=1<br />

β c 2m<br />

(2m)!<br />

[<br />

(ω3 − ω c ) 2m − ωd<br />

2m ]<br />

, (10.4.16)<br />

where the superscript c indicates that dispersion parameters are evaluat<strong>ed</strong> at the frequency<br />

ω c . This equation differs from Eq. (10.4.12) by the ω d term, which contributes<br />

only when two pumps are us<strong>ed</strong>. The main advantage of dual-pump FOPAs over singlepump<br />

FOPAs is that the ω d term can be us<strong>ed</strong> to control the phase mismatch. By properly<br />

choosing the pump wavelengths, it is possible to use this term for compensating<br />

the nonlinear phase mismatch γ(P 1 + P 2 ) such that the total phase mismatch κ is maintain<strong>ed</strong><br />

close to zero over a wide spectral range.<br />

The most commonly us<strong>ed</strong> configuration for dual-pump FOPAs employs a relatively<br />

large wavelength difference between the two pumps. At the same time, the center<br />

frequency ω c is set close to the zero-dispersion frequency ω 0 of the fiber so that the<br />

linear phase mismatch in Eq. (10.4.16) is constant over a broad range of ω 3 . To achieve<br />

a fairly wide phase-matching range, the two pump wavelengths should be locat<strong>ed</strong> on<br />

the opposite sides of the zero-dispersion wavelength in a symmetrical fashion [90].<br />

With this arrangement, κ can be r<strong>ed</strong>uc<strong>ed</strong> to nearly zero over a wide wavelength range,<br />

resulting in a gain spectrum that is nearly flat over this entire range.<br />

The prec<strong>ed</strong>ing discussion assumes that only the nondegenerate FWM process, ω 1 +<br />

ω 2 → ω 3 +ω 4 , contributes to the parametric gain. In reality, the situation is much more<br />

complicat<strong>ed</strong> for dual-pump FOPAs because the degenerate FWM process associat<strong>ed</strong><br />

with each pump occurs simultaneously. In fact, it turns out that the combination of<br />

degenerate and nondegenerate FWM processes can create up to eight other idler fields,<br />

in addition to the one at the frequency ω 4 [66]. Only four among these idlers, say,<br />

at frequencies ω 5 , ω 6 , ω 7 , and ω 8 , are important for describing the gain spectrum<br />

of a dual-pump FOPA. These four idlers are generat<strong>ed</strong> through the following FWM<br />

processes:<br />

ω 1 + ω 1 → ω 3 + ω 5 , ω 2 + ω 2 → ω 3 + ω 6 , (10.4.17)<br />

ω 1 + ω 3 → ω 2 + ω 7 , ω 2 + ω 3 → ω 1 + ω 8 . (10.4.18)<br />

In addition to these, the energy conservation requir<strong>ed</strong> for FWM is also maintain<strong>ed</strong><br />

during the following FWM processes:<br />

ω 1 + ω 1 → ω 4 + ω 7 , ω 2 + ω 2 → ω 4 + ω 8 , (10.4.19)<br />

ω 1 + ω 2 → ω 5 + ω 8 , ω 1 + ω 2 → ω 6 + ω 7 , (10.4.20)<br />

ω 1 + ω 4 → ω 2 + ω 5 , ω 1 + ω 6 → ω 2 + ω 4 . (10.4.21)

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!