Property Testing of Quantum Measurements

TL;DR

This paper introduces methods for efficiently testing properties of unknown quantum measurements, including stabilizer, k-local, and permutation-invariant measurements, with query complexity independent of system dimension.

Contribution

It develops a metric for quantum measurements and provides algorithms for property testing and measurement comparison with dimension-independent query complexity.

Findings

Efficient testing of stabilizer, k-local, and permutation-invariant measurements.

Query complexity is independent of the quantum system's dimension.

Algorithms for measuring the distance between two unknown quantum measurements.

Abstract

In this paper, we study the following question: given a black box performing some unknown quantum measurement on a multi-qudit system, how do we test whether this measurement has certain property or is far away from having this property. We call this task \textit{property testing} of quantum measurement. We first introduce a metric for quantum measurements, and show that it possesses many nice features. Then we show that, with respect to this metric, the following classes of measurements can be efficiently tested: 1. the stabilizer measurements, which play a crucial role for quantum error correction; 2. the $k$-local measurements, i.e. measurements whose outcomes depend on a subsystem of at most $k$ qudits; 3. the permutation-invariant measurements, which include those measurements used in quantum data compression, state estimation and entanglement concentration. In fact, all of them can be tested with query complexity independent of the system's dimension. Furthermore, we also present an algorithm that can test any finite set of measurements. Finally, we consider the following natural question: given two black-box measurement devices, how do we estimate their distance? We give an efficient algorithm for this task, and its query complexity is also independent of the system's dimension. As a consequence, we can easily test whether two unknown measurements are identical or very different.