arXiv:2611.00599
The exact information cost of randomized measurements in quantum metrology
Wanqing Li, Jiaxin Liu
Abstract
Randomized measurements provide state-independent readout for multiparameter quantum sensing. Although 3-designs retain a constant fraction of the quantum Fisher information (QFI) for every pure state, we show that 2-designs need not. In dimension \(d\), complete mutually unbiased bases (MUBs) retain exactly \(1/(d+1)\) of the QFI in the worst state and direction, attaining the universal lower bound for 2-designs built from orthonormal bases; symmetric informationally complete (SIC) measurements retain exactly \((1-1/\sqrt{d+1})/d\). The loss comes from aligned blind spots. Relative to a basis, pure-state changes split into population and phase parts of equal dimension, but probabilities respond only to the former. Mutual unbiasedness turns a population change for one basis into a phase change for every other. The worst case already occurs for a qubit encoded in two levels. Independent random bases instead scatter their blind spots. In sufficiently large dimensions, five Haar-random bases retain, with high probability, a dimension-independent QFI fraction for every pure state and direction. Random Clifford measurements on any number of qubits retain exactly one third. Thus blind-spot geometry, not second-order symmetry alone, controls worst-case metrological performance.
AI Summary
The paper asks whether state-independent randomized measurements forming a 2-design retain a dimension-independent fraction of the quantum Fisher information (QFI) for every pure state and tangent direction. Performance is measured by $c=\inf_{\psi,u}I(u)/J(u)$. Complete mutually unbiased bases (MUBs) have exactly $c=1/(d+1)$, attaining the universal lower bound for 2-designs assembled from orthonormal bases. General rank-one 2-designs satisfy $c\ge1/[2(d+1)]$, while every SIC has exactly $c=(1-1/\sqrt{d+1})/d$. The MUB worst case is a population change of an equal superposition of two basis vectors: one basis detects it and all unbiased bases have zero first-order response. Haar-typical states instead have worst-direction retention converging to $1/2$ under complete MUBs. Five independent Haar-random bases retain a positive dimension-independent fraction uniformly over states and directions, with probability at least $1-2^{-d}$ in sufficiently large dimensions. Global random Clifford measurements on any number of qubits retain exactly $1/3$. For odd-prime local dimension $p$, multi-qudit Clifford retention lies between $1/(4p-2)$ and $1/(p+1)$; a specified single-magic-gate construction guarantees $1/15$. Proofs use population/phase tangent-space geometry, design moment identities, small-ball estimates and covering nets, shadow-norm bounds, and a Pauli recursion. Numerical searches and estimation simulations supplement the bounds. The guarantees concern pure states; the five-basis proof gives extremely small constants, while practical constants are numerical. Whether three or four random bases suffice, and whether the multi-qudit Clifford upper bound is exact, remain open.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.