Optimal unbiased Gaussian quantum estimation and the cost of unknown squeezing
Loading original first page…
Abstract
We determine minimum-variance unbiased estimators of scalar observables from N copies of an unknown Gaussian quantum state. Real orthogonal mixing of the copies defines an expectation-complete invariant model. Its completeness makes the averaged observable optimal at every centered Gaussian state, including squeezed, multimode and pure states, among all globally unbiased integrable quantum procedures. The theorem covers bounded copy observables and number-preserving polynomials, with the unbounded case justified by absolute first-moment forms and exponential photon-number domains. Polynomial energies admit an exact finite Wick formula. For the single-mode quartic energy, we derive the optimal unbiased observable and its variance, then prove a sharp resource consequence allowing arbitrary bias: with occupation at least a, both collective quantum measurements and standard heterodyne have minimax relative risk of order min{1, 1/(Na²)} uniformly for 0 < a ≤ 1/4. Thus admitting unknown squeezing changes the low-brightness quantum copy cost from inverse-linear to inverse-quadratic occupation, matching the heterodyne order. An explicit heterodyne pair statistic attains the upper order, while a squeezed-thermal testing experiment gives the quantum lower bound. Joint location-scatter completeness extends pointwise unbiased optimality to centered targets with unknown displacement using the residual copies.
Read the full paper in the original PDF.
Edition files
- main.pdf135.9 KiB · pdf
SHA-256
989dc62bbf66bbcca58add457ca52064e990a981ef2e18cf13c6ceb94dbd3895 - Gaussian_quantum_estimation_publication.zip153.3 KiB · package
SHA-256
04576bdd1b8a43148b3bd789f2693818962876aa39d3af1feff3e604fb820233
Available editions
- Edition 1 · 2026-11-01 · public
Research details
- Subjects
- Mathematics · Quantum Information Theory · Statistics and Data Science · Statistical Theory
- Keywords
- Gaussian quantum states, unbiased estimation, statistical completeness, squeezing, minimax risk
- Manuscript date
- 2026-11-01
- Reader reuse
- none
Scope
Exact minimum-variance globally unbiased scalar estimation over the full centered Gaussian covariance family, including bounded observables and balanced Hermitian polynomials. Exact quartic risk, matched arbitrary-bias single-mode quartic quantum and heterodyne minimax orders, and centered targets with unknown displacement.
Disclosures
OpenAI Codex was used for mathematical, computational and editorial assistance in preparing this work.
Reproduction
The 14-page manuscript contains the analytic proofs. The source package supplies editable LaTeX, a portable Tectonic build script, exact-arithmetic standard-library Python controls, reader instructions, citation metadata and a checksum manifest.