arXiv:2610.14424
Exponential Quantum Advantage in Stationary Signal Sensing
Sojeong Park, Changhun Oh
Abstract
Sensing time-varying signals is essential for precision measurement and for characterizing fluctuations in physical systems. In this work, we show that quantum entanglement provides an exponential advantage in the acquisition time required to learn the temporal statistics of a stationary classical signal. The task is to estimate the characteristic function of joint displacements on selected bosonic modes within a fixed-duration window, using continuous measurements of one signal realization. We present an entanglement-assisted scheme based on two-mode squeezed probes and local Bell measurements with protected idlers, and bound its acquisition time accounting for readout noise and correlations between signal windows. For a strictly stationary pulse family, it achieves polynomial acquisition time at inverse-polynomial accuracy with logarithmic squeezing and polynomial photon number. On the other hand, we prove that any local idler-free protocols require acquisition time exponential in the number of selected temporal modes per window. This lower bound holds for both sequentially prepared and arbitrarily temporally entangled probes, allowing non-Gaussian inputs and measurements and causal classical feedback, as long as each output is measured destructively within its time bin. These results establish an exponential quantum learning advantage for a single continuously monitored stationary signal, with the sensing cost measured by total acquisition time.
AI Summary
The paper studies the acquisition time needed to estimate joint-displacement characteristic functions of selected bosonic temporal modes from one continuously monitored stationary classical signal. It compares entanglement-assisted sensing with local, idler-free protocols whose outputs are destructively measured within each time bin. Acquisition must be independent of the subsequently supplied Fourier query. Two-mode squeezed probes, protected idlers and local Bell measurements yield displacement samples with Gaussian noise variance $e^{-2r}$. An unbiased Fourier estimator achieves error $\epsilon$ with failure probability at most $\delta$ in acquisition time of order $[T\exp(2e^{-2r}\kappa n)+\tau_{\rm dep}]/(\delta\epsilon^2)$ for queries satisfying $|\beta|^2\le\kappa n$. Thus logarithmic squeezing and polynomial photon number suffice for polynomial time when accuracy is inverse-polynomial and temporal dependence is controlled. The lower bounds use a strictly stationary renewal-pulse family with hidden sinusoidal perturbations of Gaussian amplitude distributions. Learning is reduced to hypothesis testing, and Fourier-response bounds limit distinguishability per pulse. Writing $b=2\kappa+\sqrt{1+4\kappa^2}$, the resulting time bounds scale as $b^{n/2}/(n^2\epsilon^2)$ for sequentially prepared probes and $(50b/81)^{n/4}/(n^2\epsilon^2)$ for temporally entangled inputs, in units of $T$, for $\kappa\ge15/16$ and $\epsilon\le1/(24n)$. An operator second-moment argument establishes the latter extension, including non-Gaussian probes and causal classical feedback. The separation is worst-case and depends on restricted readout access, not a comparison against unrestricted idler-free sensing. Complementary modes within a packet can themselves supply an effective idler. The hard family has dependence time below $30T$ and polynomial signal power; smooth-pulse realizations cover a specified class of stable finite-response sensors.
Central claim may not hold: The abstract omits the essential requirement that acquisition be independent of the characteristic-function query: if the query is known beforehand, local squeezed-quadrature measurements can estimate its Fourier feature in polynomial acquisition time.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 29 pages, 5 figures