Abstract

In many fiber-optic applications a definite state of optical polarization is required at the end of a long fiber run. This occurs in coherent communications upon mixing with a local oscillator¹ and in fiber sensors to a void input-lead polarization noise² and polarization fading.³ In general, three approaches have been employed to solve or circumvent this problem: 1) the use of high-birefringence polarization-preserving fiber, 2) active polarization control with a birefringence device, which produces the desired state at the output, in conjunction with a feedback loop,⁴ and 3) the use of depolarized light, which will always remain depolarized, independent of fiber birefringence.^1,5 We focus here on the third approach, which has the advantage of avoiding expensive polarization-preserving fiber and does not require feedback loops.

© 1991 Optical Society of America