We present two measuring strategies reaching the Heisenberg limit for phase-shift measurements using semiclassical coherent states exclusively. We examine their performance by assuming practical experimental conditions such as losses and nonideal detectors.
The ultimate precision of phase estimation is limited by the Heisenberg scaling Δϕ0=K/N, where K∼1 is a numerical prefactor and N is the mean number of photons interacting with the phase shifting object(s). However, achieving this fundamental limit often comes at the cost of an extremely narrow high-sensitive range, rendering schemes impractical. We analyze the precision limits of phase measurements in single- and two-arm optical interferometers with input Gaussian states. We consider two detection methods: conventional homodyne measurement and non-Gaussian threshold detection that saturates the quantum Cramér-Rao bound. We characterize the performance by two complementary metrics: the peak sensitivity Δϕ0 and the width δϕ of the high-sensitivity range. We demonstrate that Heisenberg scaling is attainable in all configurations considered. However, we reveal that δϕ strongly depends on K. We derive an approximate analytic expression that describes this trade-off. We show also that the two-arm interferometer with antisymmetrically squeezed inputs exhibits exceptional performance, simultaneously achieving Heisenberg-limited sensitivity and a broad high-sensitivity range δϕ=π/2.
We analyze the quantum limit to the sensitivity of the detection of small displacements. We focus on the case of free particles and harmonic oscillators as the systems experiencing the displacement. We show that the minimum displacement detectable is proportional to the inverse of the square root of the mean value of the energy in the state experiencing the displacement (Heisenberg limit). We present a measuring scheme that reaches this limit using semiclassical states. We examine the performance of this strategy under realistic practical conditions by computing the effect of imperfections such as losses and nonunit detection efficiencies. This analysis confirms the robustness of this measuring strategy by showing that the experimental imperfections can be suitably compensated by increasing the mean energy of the input state.