Valeria Cimini - Università di Roma La Sapienza

Multiparameter quantum-enhanced phase sensing with squeezed light

Squeezed states of light are a key resource in quantum metrology, particularly for phase estimation, while remaining compatible with compact, room-temperature interferometers. In practice, however, quantum-enhanced precision can only be achieved if the measurement projection is properly optimized, making adaptive strategies essential. Current adaptive strategies suffer from two major drawbacks, on one hand they operate over at most half the phase range and on the other they depend heavily on pre-calibration of the probe parameters thus compromising the true unconditional quantum advantage of the previously developed protocols. Our approach overcomes both these limitations simultaneously by jointly inferring the nuisance parameters together with the phase, using that information in real time to set the optimal measurement projection.

Our multiparameter adaptive protocol achieves phase estimation precisions below the shot-noise limit across the full interval [0,𝜋), thereby demonstrating unconditional quantum enhancement in phase sensitivity without the necessity of pre-calibrating any parameter of the setup. This protocol therefore allows enhanced phase estimation precision, while remaining resilient to probe fluctuations, providing a framework particularly well suited to practical applications.

We experimentally demonstrate this multiparameter adaptive protocol using homodyne detection with an FPGA-based feedback adaptively updating the local oscillator phase in real time as data are acquired. The scheme achieves unconditional sub-shot-noise sensitivity across the full phase range and for different squeezing levels. Our results establish a practical pathway to self-calibrating quantum sensors based on squeezed light, as it introduces an adaptive phase estimation protocol that automatically counteracts probe and system fluctuations, applied in a relevant framework using squeezed light that can be used in different applications. This represents a step towards practical implementation scenarios in which a precise control of the experimental conditions and probe parameters is often not feasible.