Answers are grounded in this paper and cite their sources.
In this paper we utilize the idea of multiatom squeezed states to build a rigged reservoir for linear amplification or attenuation. The reservoir state is constructed such that the processed boson mode is driven by a collective reservoir operator that offers squeezed stationary noise. Then the amplifier can retain squeezing for arbitrary gain, and the amplifier added noise can be made widely different in the two quadrature components. The attenuator drives the signal toward a squeezed noise state. In contrast to the case of a multimode squeezed reservoir, we can in this case make the noise exactly zero for one component, but only at the expense of a vanishing linear gain (or loss) coefficient. The noise-over-gain ratio can, however, become arbitrarily small. The results are discussed and related to earlier work.
No takes yet. Share an insight, caveat, or question.
Dupertuis et al. (1987) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: