Raman-induced performance limits on applications of fiber parametric amplifiers.

摘要

Fiber parametric amplifiers are capable of either performing all-optical networking functions such as amplification, wavelength conversion, retiming, and reshaping, or serving as a source of entanglement for use in quantum-cryptography and quantum computing. Parametric amplifiers offer significant noise-figure benefits over semiconductor optical amplifier implementations of the above functions while commercially available erbium-doped fiber and Raman amplifiers do not provide networking functions beyond amplification. The main advantages of using parametric amplifiers as sources of entanglement are efficiency—the generation and transmission components of the system are well matched. Also, one may conveniently provide wavelength-division-multiplexed entanglement.; Parametric amplification occurs in optical fibers via a third-order nonlinear process called four-wave mixing, where one or two strong optical pump beams provide optical gain and spontaneous emission to signal and idler wavelengths simultaneously. To date, researchers have assumed that an optimally designed fiber parametric amplifier has essentially the same performance as an ideal parametric amplifier.; But careful experiments done in this lab show that unexpectedly large fluctuations are added to the signal and idler wavelengths, something not explained by a simple four-wave-mixing theory. Explaining these results is then a useful and important contribution to scientists and engineers working with fiber parametric amplifiers. Accordingly, we have developed an analytically solvable quantum theory of fiber parametric amplifiers that takes into account the noninstantaneous response of the third-order nonlinearity of optical fibers, which determines the fundamental performance limit on applications of fiber parametric amplifiers. We confirm these limits by comparison with experiment for phase-insensitive parametric amplification, for twin-beam generation, and qualitatively for twin-photon pair production. Because this quantum limit depends on the nonlinear material of the amplifier, the temperature, and how pump and signal wavelengths are chosen, one can design for reduction of the quantum limit for a particular application. In order to experimentally confirm these theories, we have also made significant developments in single-photon counting technology and in a technique called optical homodyne tomography.