Source-linked AI summary
Quantum gate verification and its application in property testing
Pei Zeng, You Zhou, Zhenhuan Liu
TL;DR
The paper develops gate verification as a benchmarking approach for target unitary gates, using local input states and measurements and analyzing both independent and adversarial settings. It derives optimal strategies and applies verification to common gates and to testing channel entanglement properties and quantum-memory robustness.
Problem
The paper addresses how to benchmark target quantum gates, including in settings where noise across trials may be correlated or channels are untrusted.
Source: A. Non-adversarial scenario · p. 3, C. Adversarial scenario · p. 5Method
It maps gate verification to Choi-state verification and analyzes local prepare-and-measure strategies without ancillaries.
Source: A. Non-adversarial scenario · p. 3, §V · p. 11Results
The paper derives optimal strategies for non-adversarial and adversarial verification and gives protocols for common gates, including single-qubit and qudit gates, Clifford gates, and multi-qubit controlled-Z and controlled-X gates.
Source: §V · p. 11Takeaways & Limitations
Gate verification can test whether channels preserve entanglement and can quantify a lower bound on quantum-memory robustness.
Source: §IV · p. 9, §V · p. 11
Abstract
from arXiv · showhide
To guarantee the normal functioning of quantum devices in different scenarios, appropriate benchmarking tool kits are quite significant. Inspired by the recent progress on quantum state verification, here we establish a general framework of verifying a target unitary gate. In both the non-adversarial and adversarial scenarios, we provide efficient methods to evaluate the performance of verification strategies for any qudit unitary gate. Furthermore, we figure out the optimal strategy and its realization with local operations. Specifically, for the commonly-used quantum gates like single qubit and qudit gates, multi-qubit Clifford gates, and multi-qubit Controlled-Z(X) gates, we provide efficient verification protocols. Besides, we discuss the application of gate verification for the detection of entanglement-preserving property of quantum channels and further quantify the robustness measure of them. We believe that the gate verification is a promising way to benchmark a large-scale quantum circuit as well as to test its property.
I. PRELIMINARIES
The Choi representation turns channel verification into verification of a corresponding state, while the average gate fidelity provides a practical measure of deviation from the target unitary. The section introduces these foundations before presenting the general gate-verification framework.
A. Choi state representation of quantum channels
- The normalized Choi state represents a channel's action on one half of a maximally entangled input and encodes all information about that channel.For a unitary channel, its Choi state is a pure maximally entangled state.Source: A. Choi state representation of… · p. 2
- State-channel duality makes verifying a quantum channel equivalent to verifying its Choi state.This lets results from quantum state verification be applied to gate verification.Source: A. Choi state representation of… · p. 2
B. Average gate fidelity and entanglement fidelity
- Average gate fidelity characterizes the difference between an ideal unitary and a noisy channel, and is directly related to the Choi states' entanglement fidelity.The paper uses entanglement infidelity, r(U, E) := 1 − F(U, E), in its analysis.Source: B. Average gate fidelity and… · p. 2
II. GENERAL FRAMEWORK OF QUANTUM GATE VERIFICATION
- The paper introduces a general quantum gate-verification framework, analyzing non-adversarial performance, optimal strategies, and adversarial verification.The adversarial setting is presented as relevant to quantum communication with untrusted channels.Source: §II · p. 2
A. Non-adversarial scenario
In the non-adversarial setting, the verifier tests a channel using randomly selected input-state and measurement pairs, then evaluates performance through the probability that a noisy channel passes. The analysis relates this probability to the required trial count and also addresses channels that vary across rounds.
A. Non-adversarial scenario
- The verifier randomly selects a pair, prepares its input state, measures the channel output with a two-outcome POVM, and accepts if all N tests pass.Passing supports the estimate that average gate fidelity is at least 1 − ϵ at significance level δ.Source: A. Non-adversarial scenario · p. 3
- The maximal pass probability for channels with entanglement infidelity at least ϵ can be evaluated by a semidefinite program, with the spectral gap characterizing the verification operator.The channel's Choi state also obeys a trace constraint not imposed in general bipartite-state verification.Source: A. Non-adversarial scenario · p. 3
- The minimum trial count is determined by the maximal pass probability for the chosen verification operator and the required significance level.The analysis also provides an upper bound on the optimal trial count.Source: A. Non-adversarial scenario · pp. 3–4
- When channels differ across rounds, the same verification schemes can bound their average infidelity.The relevant quantity is the averaged infidelity across rounds.Source: A. Non-adversarial scenario · p. 4
B. Optimal verification with pure state inputs and projective measurements
For any unitary, the paper identifies an optimal verification strategy using pure-state inputs and projective measurements. It gives practical realizations of the optimal operator and discusses a lower-setting alternative with additional trials.
B. Optimal verification with pure state inputs and projective measurements
- Any unitary gate can be optimally verified using pure-state inputs and projective measurements.The strategy for a general unitary is obtained from one for the identity channel by transforming the measurement.Source: B. Optimal verification with pure… · p. 4
- The optimal verification operator is Bell-diagonal, and its pass probability is optimal among verification strategies.The proof reduces the search to Bell-diagonal operators, for which channel and state verification coincide.Source: B. Optimal verification with pure… · p. 4
- The optimal operator can be realized with conjugate-basis tests using d + 1 mutually unbiased bases when d is a prime power, or with a weighted complex projective 2-design.These constructions implement the operator in a preparation-and-measurement manner.Source: B. Optimal verification with pure… · p. 5
- Using fewer mutually unbiased bases reduces measurement settings; with two bases, the trial count is about twice the optimal count.The bound for fewer bases may not be tight, but is described as economical.Source: B. Optimal verification with pure… · p. 5
C. Adversarial scenario
The adversarial analysis addresses correlated channels across test rounds and a separate task round whose gate fidelity must be bounded. It relates gate-verification performance to state-verification bounds and shows that homogeneous strategies are optimal within the Bell-supported class.
C. Adversarial scenario
- In the adversarial setting, noise may correlate across rounds, so the verifier tests N randomly selected rounds while reserving another round for the information-processing task.The channels may be controlled by an untrusted party that exploits correlations between rounds.Source: C. Adversarial scenario · p. 5
- State-verification parameters provide a lower bound on task-round fidelity and an upper bound on the trial count for gate verification.For general strategies, the fidelity optimization is not easy to solve analytically.Source: C. Adversarial scenario · p. 6
- For Bell-diagonal strategies, gate-verification performance reduces to the corresponding state-verification performance.The analysis then extends to Bell-supported strategies, which are no better than a homogeneous strategy for fixed N and δ.Source: C. Adversarial scenario · pp. 6–7
- Among Bell-supported strategies, a homogeneous strategy can achieve the optimal fidelity for fixed trial rounds N and significance level δ.Homogeneous strategies are a specific case of Bell-diagonal strategies.Source: C. Adversarial scenario · p. 7
- In the high-precision limit, the paper identifies an optimal homogeneous strategy and describes its realization by mixing trivial tests with the optimal conjugate-basis test.The stated realization uses the trivial test with probability p and the original test with probability 1 − p.Source: C. Adversarial scenario · p. 7
III. VERIFICATION OF SOME TYPICAL QUANTUM GATES
The section applies the general gate-verification framework to typical quantum gates, including single-qubit, multi-qubit Clifford, and other gates. It focuses on the non-adversarial scenario.
III. VERIFICATION OF SOME TYPICAL QUANTUM GATES
- The protocols cover single-qubit gates, multi-qubit Clifford gates, and other typical quantum gates, with this section focusing on the non-adversarial scenario.Source: §III · p. 7
A. Single qubit gates
The section constructs an optimal verification strategy for the qubit identity using three Pauli-basis MUBs, then extends it to other single-qubit gates and general qudit gates.
A. Single qubit gates
- Three mutually unbiased bases from the Pauli bases define the qubit identity verification strategy and its verification operator.The strategy has an optimal trial number, though the supplied passage does not provide a legible value.Source: A. Single qubit gates · p. 8
- For any single-qubit gate U, updating the verification pairs by applying U to the measurement elements gives a verification strategy.For the Z gate, the listed input-output pairs swap the ± eigenstates of X and Y while leaving the Z eigenstates unchanged.Source: A. Single qubit gates · p. 8
- The same approach verifies the non-Clifford T gate, while general qudit gates can be verified using the framework in Section II B.Source: A. Single qubit gates · p. 8
B. Clifford gates
The section gives local verification strategies for multi-qubit Clifford gates using Pauli-basis inputs and local Pauli measurements. Stabilizer-based tests can improve the spectral gap, while graph coloring can reduce measurement settings.
B. Clifford gates
- For an n-qubit Clifford circuit containing CZ gates only between neighboring qubits, two state-preparation and measurement settings suffice for efficient verification.Using all non-trivial stabilizers can increase the spectral gap, but may require more state-preparation and measurement settings.Source: B. Clifford gates · p. 8
- The local strategy realizes stabilizer projectors using product pure states in Pauli bases and measurements of Pauli operators.The specified example implements a product measurement using local Pauli measurements and classical post-processing.Source: B. Clifford gates · pp. 8–9
- The approach generalizes to verification of any Clifford gate, while the optimal strategy may be difficult to implement because its inputs and measurements can be entangled.Source: B. Clifford gates · pp. 8–9
- Any n-qubit Clifford gate can be verified with Pauli-basis input states and local Pauli measurements, with a trial-number upper bound that can be reduced using more inputs and settings.Source: B. Clifford gates · p. 9
C. Multi-qubit Control-Z and Control-X gates
For multi-qubit controlled-Z and controlled-X gates, the section maps their Choi states to hypergraph states and uses color-based verification protocols.
C. Multi-qubit Control-Z and Control-X gates
- The Choi states of C^(n−1)Z and C^(n−1)X are equivalent to hypergraph states under local Hadamard gates, enabling stabilizer-based verification.Source: C. Multi-qubit Control-Z and Control-X… · p. 9
- The relevant graph is (n+1)-colorable, giving a verification spectral gap of 1/(n+1) for both gate families under the color protocol.Source: C. Multi-qubit Control-Z and Control-X… · p. 9
- The gate verification can use local state inputs and Pauli-basis measurements because a C^kZ operation can be measured in the Z basis with post-processing.Source: C. Multi-qubit Control-Z and Control-X… · p. 9
IV. APPLICATIONS IN CHANNEL PROPERTY TESTING
The verification protocol tests whether a channel preserves entanglement, then uses verification to lower-bound quantum-memory robustness. The section gives an entanglement test using two measurement settings and relates robustness estimates to Choi-state witnesses.
A. Entanglement property detection
- The channel’s entanglement-preserving property is equivalent to entanglement of its Choi state, detected when its fidelity with the maximally entangled state is at least 1/d.The witness has nonnegative expectation on all separable states.Source: A. Entanglement property detection · p. 9
- Two measurement settings, such as X- and Z-basis states and measurements, can detect a channel’s entanglement-preserving property.The section also states that EP can be verified in a single round when d ≥ 2δ^-1 − 1.Source: A. Entanglement property detection · p. 10
B. Quantumness quantification
- Verification can lower-bound the robustness of quantum memory, an operational measure of channel quantumness.The generalized channel robustness has operational meanings including classical simulation cost and advantage in state-discrimination games.Source: B. Quantumness quantification · pp. 10–11
- The Choi-state robustness bound is conservative because its minimization omits the maximally-mixed-first-subsystem constraint on valid channel Choi states.The section notes that using fewer mutually unbiased bases can reduce measurement effort.Source: B. Quantumness quantification · p. 11
- To certify R(E) ≥ r, the entanglement infidelity must satisfy ϵ ≤ (d − r − 1)/d.The channel-robustness estimate is lower-bounded using generalized entanglement robustness of the channel’s Choi state and a witness.Source: B. Quantumness quantification · p. 11
V. CONCLUSION AND OUTLOOK
The paper develops local-input, local-measurement gate verification without extra ancillas, with optimal strategies for non-adversarial and adversarial settings. It applies these methods to common gates and channel-property tests, while identifying robustness to preparation, measurement, and few-round failure as future work.
V. CONCLUSION AND OUTLOOK
- Non-adversarial verification performance is computable by semidefinite programming, and a uniformly mixed strategy is optimal and can be realized with d+1 MUBs for prime-power d or a quantum state 2-design.The performance of all Bell-diagonal strategies can also be evaluated exactly.Source: §V · p. 11
- For adversarial verification, state parameters bound generally hard-to-solve fidelity and trial-count quantities, and homogeneous strategies attain the optimal fidelity among Bell-supported strategies.The optimality claim is for fixed trial rounds and significance level.Source: §V · p. 11
- The analysis covers local verification of single-qubit and qudit gates, multi-qubit Clifford gates, and multi-qubit controlled-Z and controlled-X gates.The strategies use local state inputs and measurements without extra ancillas.Source: §V · p. 11
- Gate verification is applied to testing entanglement preservation and quantifying quantum-memory robustness, and the authors suggest extending it to other channel properties.Examples proposed for extension include coherence-generating power and magic.Source: §V · p. 11
- The authors identify robustness to state-preparation and measurement errors and to failure tests in a few rounds as future improvements.They suggest combining channel verification with randomized benchmarking, robust tomographic information extraction, or gate-set tomography for preparation and measurement errors.Source: §V · p. 11
- The paper reports a better trial-number performance than Ref. [54] for the multi-qubit Clifford gate comparison.The comparison is described for the gate in Eq. (55).Source: §V · p. 12
Appendix A: Heisenberg-Weyl Operators and generalized qudit Bell states
Appendix A introduces Heisenberg-Weyl operators and generalized qudit Bell-state notation. Appendix B proves that twirling produces a Bell-diagonal strategy without increasing its passing probability.
Appendix A: Heisenberg-Weyl Operators and generalized qudit Bell states
- The Heisenberg-Weyl group generalizes the Pauli group; the cited passages define a qudit operator W(u, v) and denote qudit Bell-state projectors as Φu,v := |Φu,v⟩⟨Φu,v|.Source: Appendix A · p. 14
Appendix B: Proof of Lemma 2
- Twirling removes off-diagonal Bell-basis terms, leaving a Bell-diagonal verification operator.The proof uses the commuting relation and invariance of the maximally entangled state under W* ⊗ W.Source: Appendix B · pp. 14–15
- Twirling does not increase the passing probability because it mixes the transformed verification operators with equal probability.The proof applies the stated observation and lemma to this mixture.Source: Appendix B · p. 15
Appendix C: Proof of Lemma 3
The proof reduces the adversarial optimization to Choi states diagonal in the strategy’s eigenbasis and organizes feasible mixtures through permutation-invariant symmetric Bell states. It then shows that symmetric Bell strategies for the twirled strategy can be realized by the original strategy with the same relevant parameters.
Appendix C: Proof of Lemma 3
- Permutation-invariant symmetrization across rounds yields symmetric Bell states when the constituent basis states are maximally entangled.The proof denotes this subset by ΘS(Ψj,l, N).Source: Appendix C · p. 15
- Because pE and fE depend only on diagonal Choi-state elements in the strategy basis, the proof assumes the Choi state is diagonal there.Source: Appendix C · p. 15
- Legal mixture coefficients are nonnegative, normalized, and constrained by the Choi-state requirement; their feasible set is convex.The legal set is denoted C(Ψk) and is determined by the symmetric-state choice.Source: Appendix C · p. 15
- Every symmetric Bell strategy for Ω′ can be realized on Ω with the same (pE(c), fE(c)), supporting F(N, δ, Ω′) ≥ F(N, δ, Ω).The proof relies on the same feasible coefficient region and matching ηk and ζk values for the symmetric Bell basis.Source: Appendix C · p. 16