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.

10.5. Polarization Effects 403<br />

following set of four equations [104]:<br />

d|A 1 〉<br />

dz<br />

d|A 2 〉<br />

dz<br />

d|A 3 〉<br />

dz<br />

d|A 4 〉<br />

dz<br />

= 2iγ (<br />

)<br />

〈A 1 |A 1 〉 + 〈A 2 |A 2 〉 + 1<br />

3<br />

2 |A∗ 1〉〈A ∗ 1| + |A 2 〉〈A 2 | + |A ∗ 2〉〈A ∗ 2| |A 1 〉, (10.5.6)<br />

= 2iγ<br />

3<br />

= 2iγ<br />

3<br />

(<br />

)<br />

〈A 1 |A 1 〉 + 〈A 2 |A 2 〉 +<br />

2 1|A∗ 2〉〈A ∗ 2| + |A 1 〉〈A 1 | + |A ∗ 1〉〈A ∗ 1| |A 2 〉, (10.5.7)<br />

(<br />

)<br />

〈A 1 |A 1 〉 + |A 1 〉〈A 1 | + |A ∗ 1〉〈A ∗ 1| + 〈A 2 |A 2 〉 + |A 2 〉〈A 2 | + |A ∗ 2〉〈A ∗ 2| |A 3 〉<br />

+ 2iγ (<br />

)<br />

|A 2 〉〈A ∗<br />

3<br />

1| + |A 1 〉〈A ∗ 2| + 〈A ∗ 1|A 2 〉 |A ∗ 4〉e −iΔkz , (10.5.8)<br />

= 2iγ (<br />

)<br />

〈A 1 |A 1 〉 + |A 1 〉〈A 1 | + |A ∗<br />

3<br />

1〉〈A ∗ 1| + 〈A 2 |A 2 〉 + |A 2 〉〈A 2 | + |A ∗ 2〉〈A ∗ 2| |A 4 〉<br />

+ 2iγ (<br />

)<br />

|A 2 〉〈A ∗<br />

3<br />

1| + |A 1 〉〈A ∗ 2| + 〈A ∗ 1|A 2 〉 |A ∗ 3〉e −iΔkz . (10.5.9)<br />

In the case of single-pump configuration, only the first pump equation should be retain<strong>ed</strong><br />

because |A 2 〉 = 0. Also, one should replace |A 2 〉 with |A 1 〉 and 2γ with γ in the<br />

signal and idler equations.<br />

10.5.2 Polarization Dependence of Parametric Gain<br />

The vector FWM equations, Eqs. (10.5.6) through (10.5.9), describe FWM in the general<br />

case in which the two pumps and the signal are launch<strong>ed</strong> into an optical fiber with<br />

arbitrary SOPs. Their complexity requires a numerical approach in general. To investigate<br />

the relationship between the FWM efficiency and pump SOPs, we ignore the SPM<br />

and XPM terms temporarily and focus on the selection rules that govern the creation<br />

of idler photons. The SPM and XPM terms affect only the phase-matching condition<br />

and their impact is consider<strong>ed</strong> later in this section.<br />

Physically, the polarization dependence of FWM stems from the requirement of<br />

angular momentum conservation among the four interacting photons in an isotropic<br />

m<strong>ed</strong>ium. This requirement can be describ<strong>ed</strong> most simply in a basis in which ↑ and<br />

↓ denote the left and right circular polarization states and represent photons with an<br />

intrinsic angular momentum (spin) of +¯h and −¯h, respectively [105]. To describe<br />

FWM among arbitrarily polariz<strong>ed</strong> optical fields, we decompose the Jones vector of<br />

each field as<br />

|A j 〉 = U j |↑〉+ D j |↓〉, (10.5.10)<br />

where U j and D j represent the field amplitudes in the spin-up and spin-down states,<br />

respectively, for the jth wave ( j = 1 to 4). Using this expansion, it follows from Eq.<br />

(10.5.9) that the creation of idler photons in the two orthogonal states is govern<strong>ed</strong> by<br />

the following two equations (assuming perfect phase matching):<br />

dU 4<br />

dz<br />

dD 4<br />

dz<br />

= 4iγ<br />

3 [U 1U 2 U ∗<br />

3 +(U 1 D 2 + D 1 U 2 )D ∗ 3 ], (10.5.11)<br />

= 4iγ<br />

3 [D 1D 2 D ∗ 3 +(U 1 D 2 + D 1 U 2 )U ∗<br />

3 ]. (10.5.12)

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

Saved successfully!

Ooh no, something went wrong!