The Metrology Problem

Quantum metrology asks a deceptively simple question: given a physical quantity we cannot observe directly — a magnetic field, an optical phase, a frequency, an elapsed time — how precisely can we estimate it, and how do the laws of quantum mechanics set that limit? This module treats sensing as a parameter-estimation problem and shows how entanglement and quantum coherence let well-designed probes outperform any classical strategy.

The estimation pipeline

Every quantum sensing experiment follows the same four-stage loop:

Preparation. Initialise a probe in a chosen quantum state $|\psi_0\rangle$ .
Encoding. Let the probe interact with the system for a time, so the unknown parameter $\theta$ becomes imprinted on the state through a unitary $U(\theta) = e^{-i\theta G}$ , where $G$ is the generator coupling the probe to the parameter.
Measurement. Read out the probe with a chosen positive operator-valued measure (POVM), producing outcomes whose statistics depend on $\theta$ .
Estimation. Feed the outcomes to a classical estimator $\hat\theta$ that returns a best guess of the true value.

The whole point is to choose the probe state and the measurement so that small changes in $\theta$ produce large, distinguishable changes in the outcome statistics.

What "precision" means

Repeat the experiment and the estimator $\hat\theta$ fluctuates. For an unbiased estimator ( $\langle\hat\theta\rangle = \theta$ ) the figure of merit is the variance

(\Delta\theta)^2 = \big\langle (\hat\theta - \theta)^2 \big\rangle ,

or equivalently its square root, the standard deviation $\Delta\theta$ . Smaller variance means a sharper estimate. A central question of metrology is how $\Delta\theta$ scales with the resources we spend — typically the number $N$ of probes (atoms, photons) or the total interrogation time.

Why quantum mechanics matters

Classically, averaging $N$ independent noisy measurements reduces the standard deviation by a factor of $\sqrt{N}$ . Quantum mechanics offers two distinct opportunities to do better:

Coherence lets a probe accumulate phase undisturbed over a long interrogation time, sharpening the signal before measurement noise enters.
Entanglement correlates $N$ probes so that they accumulate phase $N$ times faster than independent probes, in principle replacing the $1/\sqrt{N}$ improvement with a $1/N$ improvement.

The first effect underpins the standard quantum limit; the second can reach the Heisenberg limit. The rest of this module makes both limits precise, derives them from first principles, and shows how to realise the underlying probe states on the simulator.

The roadmap

We will build the toolkit in stages: the standard quantum limit and its $1/\sqrt{N}$ scaling; the classical and quantum Fisher information that quantify how much a measurement learns about $\theta$ ; the Cramér–Rao bound that turns Fisher information into a hard precision floor; Ramsey interferometry as the workhorse phase-estimation protocol; entangled GHZ probes that beat the standard limit; the Heisenberg limit they approach; and the noise processes that ultimately cap real devices. By the end you will be able to design a small entangled probe and demonstrate, on the simulator, a measurable precision advantage over the classical baseline.