[Paper Review] Decoherence-enhanced measurements

Daniel Braun, John Martin|ArXiv.org|Feb 7, 2009

Quantum Information and Cryptography26 references3 citations

TL;DR

This paper proposes a novel measurement principle called Decoherence-Enhanced Measurements (DEM), where decoherence—typically a nuisance in quantum systems—is harnessed as a sensitive probe to achieve Heisenberg-limited precision. By preparing a product state in a decoherence-free subspace (DFS), the method achieves a signal-to-noise ratio scaling as $\sim N$, enabling high-precision measurement of small changes in cavity length without requiring highly entangled states.

ABSTRACT

Quantum-enhanced measurements use highly non-classical quantum states in order to enhance the precision of the measurement of classical quantities, like the length of an optical cavity. The major goal is to beat the standard quantum limit (SQL), i.e. a precision of order $1/\sqrt{N}$, where $N$ is the number of quantum resources (e.g. the number of photons or atoms used), and to achieve a scaling $1/N$, known as the Heisenberg limit. Doing so would have tremendous impact in many areas, but so far only three experiments have demonstrated a slight improvement over the SQL. The required quantum states are generally difficult to produce, and very prone to decoherence. Here we show that decoherence itself may be used as an extremely sensitive probe of system properties. This should allow for a new measurement principle with the potential to achieve the Heisenberg limit without the need to produce highly entangled states.

Motivation & Objective

• To develop a new quantum measurement principle that leverages decoherence rather than suppressing it.
• To achieve Heisenberg-limited sensitivity ($1/N$ scaling) without requiring the preparation of highly entangled macroscopic states.
• To demonstrate that collective decoherence rates in a decoherence-free subspace are highly sensitive to small perturbations in system parameters, such as cavity length.
• To identify experimentally feasible initial states—specifically product states and Schrödinger cat states within a DFS—that enable quadratic scaling of the measurement signal with $N$.
• To show that the method remains robust under imperfect state preparation and environmental noise, maintaining $N^2$ scaling of the signal-to-noise ratio.

Proposed method

• The system consists of $N$ two-level atoms in a cavity with a semi-transparent mirror, coupled to a single electromagnetic mode, where collective decoherence is governed by a Lindblad master equation with $\mathcal{L}_c$.
• The initial state is prepared in a decoherence-free subspace (DFS) where $J_-|\psi_0\rangle = 0$, ensuring no collective emission under ideal symmetry.
• A small perturbation $\delta L$ in cavity length alters the coupling strengths $g_i$, breaking the symmetry and inducing a measurable collective photon emission rate $\alpha \propto N^2$.
• The method uses the rate of collective photon emission through the mirror as a probe of the perturbation, with signal-to-noise ratio scaling as $\sim \sqrt{\gamma \Delta t} |\delta L / L| N$.
• Numerical simulations and exact diagonalization of the Lindblad operator $\mathcal{L}_c$ are used to compute $\alpha = -\langle \dot{J}_z(0) \rangle_c$, showing $\alpha \propto N^2$ for symmetric states.
• The approach is tested on both product states in the DFS and macroscopic Schrödinger cat states, both of which exhibit $N^2$ scaling of the emission rate.

Experimental results

Research questions

• RQ1 Can decoherence be used as a resource rather than a detriment in quantum metrology?
• RQ2 Can Heisenberg-limited sensitivity be achieved without preparing highly entangled states?
• RQ3 How does the collective emission rate scale with $N$ when the system is prepared in a DFS and subjected to a small perturbation?
• RQ4 Is the $N^2$ scaling of the signal robust under imperfect state preparation or finite-time relaxation?
• RQ5 What types of initial states in the DFS yield maximal sensitivity to small changes in cavity length?

Key findings

• The collective photon emission rate $\alpha$ scales as $N^2$ for both product states and Schrödinger cat states in the DFS, enabling Heisenberg-limited sensitivity.
• For $N=2$, the exact analytical result gives $\alpha = 1/2$ in units of $\gamma |\tilde{G}_1 - \tilde{G}_2|^2$, confirming the $N^2$ scaling.
• Numerical simulations show that $-B$ in the fit $\alpha = A + B\sin(2\delta) + C\cos(4\delta)$ scales as $N^{2.4}$ for $N=8$ to $18$, suggesting $N^2$ scaling dominates for large $N$.
• The signal-to-noise ratio scales as $\sim \sqrt{\gamma \Delta t} |\delta L / L| N$, achieving the Heisenberg limit without requiring entangled states.
• Even with imperfect state preparation, the method retains $N^2$ scaling of the signal, provided measurement is delayed until after the initial emission phase.
• Randomly chosen states within the DFS yield only $\langle n_{\text{ph}} \rangle \propto N$, but the product state (5) and cat states achieve $\langle n_{\text{ph}} \rangle \propto N^2$, highlighting the importance of state selection.