Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2610.21596

Time-optimal robust quantum control for partially known errors

Llorenç Balada Gaggioli

quant-phSubmitted Oct 15, 2026• Announced Oct 16, 2026• 0 citations
arXivHTMLPDF

Abstract

Robust quantum control usually assumes one of two extremes: the error is of a known form and only its strength is unknown, or nothing at all is known about it. In practice one often sits in between. The size of a static error is unknown, but its class is known, for example, it only shifts energy levels, or it acts on each subsystem separately. We ask how fast a control can cancel, to first order, every error in such a class. A time-independent Hamiltonian can do this if and only if all errors in the class have zero diagonal in a common basis. In that case an equally spaced energy ladder in this basis does it after one period. A lower bound that holds for any time-dependent control with bounded drive power shows that this simple constant control is optimal for several natural classes, including unknown level shifts and unknown local errors on two subsystems. Completely unknown errors need a time-dependent control, and there, knowing the error class shortens the minimum time by a factor $\sqrt{d^2+1}$. So partial knowledge of the error changes not only how long an optimal robust control takes, but also its form.

AI Summary

The paper asks for the minimum duration of an identity operation that cancels, to first order, every static error in a specified linear operator space. Controls may be arbitrarily time dependent, subject to the total-power constraint $2\|H(t)\|_F^2\le\Omega^2$, and must return to the identity up to global phase. A constant robust control exists exactly when all errors have zero diagonal in a common orthonormal basis. An equally spaced energy ladder in that basis provides a construction. A lower bound for arbitrary controls follows by treating propagator matrix elements as temporal signals: robustness bounds their Gram-matrix eigenvalues, while Fourier-mode occupation bounds the integrated control energy. If $\Gamma\le gI$ and $L=d^2/g$ is integral, then $\Omega T\ge2\pi\sqrt{d(L^2-1)/6}$. For unknown level shifts, arbitrary local errors on two equal-dimensional subsystems, and qubit errors confined to a plane, constant ladders attain $\Omega T_{\min}=2\pi\sqrt{d(d^2-1)/6}$. For completely unknown errors, an existing continuous unitary-design construction attains $2\pi\sqrt{d(d^4-1)/6}$, establishing an exact duration ratio $\sqrt{d^2+1}$. For two qubits, a constant anisotropic $XX$–$ZZ$ coupling cancels local errors $\sqrt2$ faster than optimal local-only controls under the same power constraint. Results concern first-order cancellation of static errors and robust identities, not general gates. Numerical constants depend on the chosen norm and unrestricted coupling availability. General error classes, unequal subsystem dimensions, hardware-constrained controls, and higher-order time optima remain unresolved. Reversing a constant control after one period cancels second-order error, without an optimality claim.

Little sign of AI-generated textStrong assumptions

Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.

Comments
11 pages total, 7 pages main text, 2 figures

Related papers

Fundamental Costs of Noise-Robust Quantum Control: Speed Limits and Complexity

Junkai Zeng, Xiu–Hao Deng

Noise is ubiquitous in quantum systems and is a major obstacle for the advancement of quantum information science. Noise-robust quantum control achieves high-fidelity operations by engineering the evolution path so that first-order noise contributions cancel at the final time. Such dynamical error correction typically incurs a time overhead beyond standard quantum speed limits. We derive general lower bounds on control complexity that quantify this overhead for quasi-static coherent noise under bounded control amplitude. For a single noise source, we prove a universal time lower bound for first-order robustness and give a constructive scheme that implements any target gate robustly in time 4T plus a constant time. For robustness against an entire noise space, we show dimension lower bounds on the number M of segments in any mixed-unitary schedule from two mechanism: (i) a coherent dimension bound when the error subspace contains an irreducible block isomorphic to su(q), and (ii) a projection dimension bound when the noise space contains the trace-zero span of orthogonal projectors. Under bounded speed, these bounds on number of segments imply time lower bounds. With only local controls robust against noise space defined on a graph, we obtain a graph-orthogonality time bound scales linear with graph chromatic number. We illustrate the bounds through examples. Collectively, these results establish quantitative limitations on the feasibility of first-order noise-resilient operations.

arXiv:2510.07183•Oct 8, 2025•3 citations

Quantum Robust Control using Geometric Optimal Control Theory

Francesca Albertini, Domenico D'Alessandro

In this paper, we demonstrate an approach to quantum robust control based on the tools of geometric optimal control. The central objects of interest are the sensitivity functions defined as the coefficients in the Taylor expansion of the trajectory with respect to the (unknown, small) parameters which describe the deviation of the actual model from nominal one. In terms of these quantities, we formalize an optimal control problem where one searches for the optimal nominal trajectory which minimizes the size of the sensitivity while taking into account other aspects of the control design such as the energy of the control field. We consider in detail the case of a single qubit with a dephasing Hamiltonian term, and the optimal control problem of obtaining a state transfer by minimizing the weighted sum of the energy of the controlling field and the first order sensitivity. At the limit of a very large weight on the sensitivity, we obtain the optimal control which zeros the sensitivity and minimizes the control field energy. This problem has a rich mathematical structure which enables its solution in terms of elliptic integrals. For this problem, we obtain an explicit solution which is particularly simple and also smooth, avoiding discontinuities which are present in other approaches. We extend the results to the robust control of two quantum bits minimizing cross-talk contamination, as we show that such a problem decouples in two independent one qubit problems.

quant-pharXiv:2603.27894•Mar 31, 2026•2 citations

Universally Robust Quantum Control

Pablo M. Poggi, Gabriele De Chiara, Steve Campbell +1 more

We study the robustness of the evolution of a quantum system against small uncontrolled variations in parameters in the Hamiltonian. We show that the fidelity susceptibility, which quantifies the perturbative error to leading order, can be expressed in superoperator form and use this to derive control pulses that are robust to any class of systematic unknown errors. The proposed optimal control protocol is equivalent to searching for a sequence of unitaries that mimics the first-order moments of the Haar distribution, i.e., it constitutes a 1-design. We highlight the power of our results for error-resistant single- and two-qubit gates.

arXiv:2309.14437•Physical Review Letters•May 7, 2024•42 citations

Learning Arbitrary Lindbladians from Time Evolution

Zhili Chen, Zhan Yu

We study the problem of learning arbitrary $n$-qubit Lindbladians, the generators of Markovian open-system dynamics, from their time evolution. We give the first known efficient algorithm that learns arbitrary Lindbladians without any structural assumptions. For an $n$-qubit Lindbladian of dynamical strength at most $Λ$, the algorithm estimates every coefficient to error $\varepsilon$ using $\widetilde O(Λ^2/\varepsilon^2)$ experiments and $\widetilde O(Λ/\varepsilon^2)$ total evolution time, with classical running time polynomial in $n$ and $Λ/\varepsilon$. Both experimental costs match known lower bounds up to polylogarithmic factors. The algorithm identifies candidate labels through product Pauli experiments and estimates their coefficients using random stabilizer preparations and Clifford measurements. It requires no ancillas or control during the unknown evolution. For constant $k$, our algorithm learns generic $k$-local Lindbladians of local dynamical strength at most $Λ_{\mathrm{loc}}$ using $\widetilde O(n^2Λ_{\mathrm{loc}}^2/\varepsilon^2)$ experiments and $\widetilde O(nΛ_{\mathrm{loc}}/\varepsilon^2)$ total evolution time, improving the upper bounds of prior works in terms of the dependence on $n$. We further improve the lower bounds of Arad, Chen, Guo, Rebentrost, and Yu (SODA 2027) using a family of $3$-local Lindbladians. They already hold when the support is known, and only one Hamiltonian coefficient is requested, even for adaptive protocols with arbitrary ancillas and quantum memory. These lower bounds establish that our algorithm is optimal up to polylogarithmic factors in both experiment count and total evolution time for every fixed $k \geq 3$.

quant-pharXiv:2607.28610•Jul 31, 2026•5 citations

A Fundamental Bound for Robust Quantum Gate Control

Robert L. Kosut, Daniel A. Lidar, Herschel Rabitz

We derive a universal performance limit for coherent quantum control in the presence of modeled and unmodeled uncertainties. For any target unitary $W$ that is implementable in the absence of error, we prove that the worst-case (and hence the average) gate fidelity obeys the lower bound $F \ge \Flb\bigl(\tf \Omeff\bigr)$, where $\tf$ is the gate duration and $\Omeff$ is a single frequency-like measure that aggregates \emph{all} bounded uncertainty sources, e.g., coherent control imperfections, unknown couplings, and residual environment interactions, without assuming an initially factorizable system-bath state or a completely positive map. The bound is obtained by combining an interaction-picture averaging method with a Bellman-Gronwall inequality and holds for any finite-norm Hamiltonian decomposition. Hence it applies equally to qubits, multi-level qudits, and ancilla-assisted operations. Because $\Flb$ depends only on the dimensionless product $\tf\Omeff$, it yields a device-independent metric that certifies whether a given hardware platform can, in principle, reach a specified fault-tolerance threshold, and also sets a quantitative target for robust-control synthesis and system identification.

arXiv:2507.01215•Jul 1, 2025•5 citations

Hamiltonian Learning and Certification via Eigenphase Engineering

Myeongjin Shin, Yu Tong

We study Hamiltonian learning for $k$-local Hamiltonians at large step size, using only forward time evolution at integer multiples of a fixed time interval. For this task, our new technique, which we call eigenphase engineering, achieves Heisenberg-limited learning with a step size that is optimal up to logarithmic factors. For an unknown $s$-sparse Hamiltonian with fixed locality $k$ and $\|H\|_{\mathrm{op}}\leqΛ$, our algorithm recovers its Pauli coefficients to $\ell_2$ error $\varepsilon$ in total evolution time $\widetilde{O}(s/\varepsilon)$, matching the best known scaling while allowing a near-maximal step size $\widetildeΘ(1/Λ)$. As an independent contribution, we strengthen the evolution-time lower bound to $Ω_k(s^{1-1/(2k)}/\varepsilon)$, placing our algorithm within a factor $\widetilde{O}(s^{1/(2k)})$ of being optimal. Eigenphase engineering departs from approaches based on dynamical approximation or iteratively canceling the unknown Hamiltonian. It encodes energy expectation differences in the perturbative response of eigenphases of the actual controlled evolution, then extracts them through phase estimation and high-order extrapolation. This avoids the need for polynomially shorter evolution steps as the target precision improves. A version using only single-qubit operations retains the same step size with a $\sqrt n$ overhead in $\ell_2$ learning time for $n$ qubits. The method also enables tolerant Hamiltonian certification and estimation of a specified Pauli coefficient, the latter without locality or sparsity assumptions. These results establish eigenphase engineering as a versatile approach to extracting Hamiltonian information with infrequent control.

quant-pharXiv:2609.26596•Sep 23, 2026•3 citations
Robust quantum control usually assumes one of two extremes: the error is of a known form and only its strength is unknown, or nothing at all is known about it. In practice one often sits in between. The size of a static error is unknown, but its class is known, for example, it only shifts energy levels, or it acts on each subsystem separately. We ask how fast a control can cancel, to first order, every error in such a class. A time-independent Hamiltonian can do this if and only if all errors in the class have zero diagonal in a common basis. In that case an equally spaced energy ladder in this basis does it after one period. A lower bound that holds for any time-dependent control with bounded drive power shows that this simple constant control is optimal for several natural classes, including unknown level shifts and unknown local errors on two subsystems. Completely unknown errors need a time-dependent control, and there, knowing the error class shortens the minimum time by a factor d2+1\sqrt{d^2+1}d2+1​. So partial knowledge of the error changes not only how long an optimal robust control takes, but also its form.
The paper asks for the minimum duration of an identity operation that cancels, to first order, every static error in a specified linear operator space. Controls may be arbitrarily time dependent, subject to the total-power constraint 2∥H(t)∥F2≤Ω22\|H(t)\|_F^2\le\Omega^22∥H(t)∥F2​≤Ω2, and must return to the identity up to global phase.
A constant robust control exists exactly when all errors have zero diagonal in a common orthonormal basis. An equally spaced energy ladder in that basis provides a construction. A lower bound for arbitrary controls follows by treating propagator matrix elements as temporal signals: robustness bounds their Gram-matrix eigenvalues, while Fourier-mode occupation bounds the integrated control energy. If Γ≤gI\Gamma\le gIΓ≤gI and L=d2/gL=d^2/gL=d2/g is integral, then ΩT≥2πd(L2−1)/6\Omega T\ge2\pi\sqrt{d(L^2-1)/6}ΩT≥2πd(L2−1)/6​.
For unknown level shifts, arbitrary local errors on two equal-dimensional subsystems, and qubit errors confined to a plane, constant ladders attain ΩTmin⁡=2πd(d2−1)/6\Omega T_{\min}=2\pi\sqrt{d(d^2-1)/6}ΩTmin​=2πd(d2−1)/6​. For completely unknown errors, an existing continuous unitary-design construction attains 2πd(d4−1)/62\pi\sqrt{d(d^4-1)/6}2πd(d4−1)/6​, establishing an exact duration ratio d2+1\sqrt{d^2+1}d2+1​. For two qubits, a constant anisotropic XXXXXX–ZZZZZZ coupling cancels local errors 2\sqrt22​ faster than optimal local-only controls under the same power constraint.
Results concern first-order cancellation of static errors and robust identities, not general gates. Numerical constants depend on the chosen norm and unrestricted coupling availability. General error classes, unequal subsystem dimensions, hardware-constrained controls, and higher-order time optima remain unresolved. Reversing a constant control after one period cancels second-order error, without an optimality claim.
Noise is ubiquitous in quantum systems and is a major obstacle for the advancement of quantum information science. Noise-robust quantum control achieves high-fidelity operations by engineering the evolution path so that first-order noise contributions cancel at the final time. Such dynamical error correction typically incurs a time overhead beyond standard quantum speed limits. We derive general lower bounds on control complexity that quantify this overhead for quasi-static coherent noise under bounded control amplitude. For a single noise source, we prove a universal time lower bound for first-order robustness and give a constructive scheme that implements any target gate robustly in time 4T plus a constant time. For robustness against an entire noise space, we show dimension lower bounds on the number M of segments in any mixed-unitary schedule from two mechanism: (i) a coherent dimension bound when the error subspace contains an irreducible block isomorphic to su(q), and (ii) a projection dimension bound when the noise space contains the trace-zero span of orthogonal projectors. Under bounded speed, these bounds on number of segments imply time lower bounds. With only local controls robust against noise space defined on a graph, we obtain a graph-orthogonality time bound scales linear with graph chromatic number. We illustrate the bounds through examples. Collectively, these results establish quantitative limitations on the feasibility of first-order noise-resilient operations.
In this paper, we demonstrate an approach to quantum robust control based on the tools of geometric optimal control. The central objects of interest are the sensitivity functions defined as the coefficients in the Taylor expansion of the trajectory with respect to the (unknown, small) parameters which describe the deviation of the actual model from nominal one. In terms of these quantities, we formalize an optimal control problem where one searches for the optimal nominal trajectory which minimizes the size of the sensitivity while taking into account other aspects of the control design such as the energy of the control field. We consider in detail the case of a single qubit with a dephasing Hamiltonian term, and the optimal control problem of obtaining a state transfer by minimizing the weighted sum of the energy of the controlling field and the first order sensitivity. At the limit of a very large weight on the sensitivity, we obtain the optimal control which zeros the sensitivity and minimizes the control field energy. This problem has a rich mathematical structure which enables its solution in terms of elliptic integrals. For this problem, we obtain an explicit solution which is particularly simple and also smooth, avoiding discontinuities which are present in other approaches. We extend the results to the robust control of two quantum bits minimizing cross-talk contamination, as we show that such a problem decouples in two independent one qubit problems.
We study the robustness of the evolution of a quantum system against small uncontrolled variations in parameters in the Hamiltonian. We show that the fidelity susceptibility, which quantifies the perturbative error to leading order, can be expressed in superoperator form and use this to derive control pulses that are robust to any class of systematic unknown errors. The proposed optimal control protocol is equivalent to searching for a sequence of unitaries that mimics the first-order moments of the Haar distribution, i.e., it constitutes a 1-design. We highlight the power of our results for error-resistant single- and two-qubit gates.
We study the problem of learning arbitrary nnn-qubit Lindbladians, the generators of Markovian open-system dynamics, from their time evolution. We give the first known efficient algorithm that learns arbitrary Lindbladians without any structural assumptions. For an nnn-qubit Lindbladian of dynamical strength at most ΛΛΛ, the algorithm estimates every coefficient to error ε\varepsilonε using O~(Λ2/ε2)\widetilde O(Λ^2/\varepsilon^2)O(Λ2/ε2) experiments and O~(Λ/ε2)\widetilde O(Λ/\varepsilon^2)O(Λ/ε2) total evolution time, with classical running time polynomial in nnn and Λ/εΛ/\varepsilonΛ/ε. Both experimental costs match known lower bounds up to polylogarithmic factors. The algorithm identifies candidate labels through product Pauli experiments and estimates their coefficients using random stabilizer preparations and Clifford measurements. It requires no ancillas or control during the unknown evolution. For constant kkk, our algorithm learns generic kkk-local Lindbladians of local dynamical strength at most ΛlocΛ_{\mathrm{loc}}Λloc​ using O~(n2Λloc2/ε2)\widetilde O(n^2Λ_{\mathrm{loc}}^2/\varepsilon^2)O(n2Λloc2​/ε2) experiments and O~(nΛloc/ε2)\widetilde O(nΛ_{\mathrm{loc}}/\varepsilon^2)O(nΛloc​/ε2) total evolution time, improving the upper bounds of prior works in terms of the dependence on nnn. We further improve the lower bounds of Arad, Chen, Guo, Rebentrost, and Yu (SODA 2027) using a family of 333-local Lindbladians. They already hold when the support is known, and only one Hamiltonian coefficient is requested, even for adaptive protocols with arbitrary ancillas and quantum memory. These lower bounds establish that our algorithm is optimal up to polylogarithmic factors in both experiment count and total evolution time for every fixed k≥3k \geq 3k≥3.
We derive a universal performance limit for coherent quantum control in the presence of modeled and unmodeled uncertainties. For any target unitary WWW that is implementable in the absence of error, we prove that the worst-case (and hence the average) gate fidelity obeys the lower bound $F \ge \Flb\bigl(\tf \Omeff\bigr)$, where $\tf$ is the gate duration and $\Omeff$ is a single frequency-like measure that aggregates \emph{all} bounded uncertainty sources, e.g., coherent control imperfections, unknown couplings, and residual environment interactions, without assuming an initially factorizable system-bath state or a completely positive map. The bound is obtained by combining an interaction-picture averaging method with a Bellman-Gronwall inequality and holds for any finite-norm Hamiltonian decomposition. Hence it applies equally to qubits, multi-level qudits, and ancilla-assisted operations. Because $\Flb$ depends only on the dimensionless product $\tf\Omeff$, it yields a device-independent metric that certifies whether a given hardware platform can, in principle, reach a specified fault-tolerance threshold, and also sets a quantitative target for robust-control synthesis and system identification.
We study Hamiltonian learning for kkk-local Hamiltonians at large step size, using only forward time evolution at integer multiples of a fixed time interval. For this task, our new technique, which we call eigenphase engineering, achieves Heisenberg-limited learning with a step size that is optimal up to logarithmic factors. For an unknown sss-sparse Hamiltonian with fixed locality kkk and ∥H∥op≤Λ\|H\|_{\mathrm{op}}\leqΛ∥H∥op​≤Λ, our algorithm recovers its Pauli coefficients to ℓ2\ell_2ℓ2​ error ε\varepsilonε in total evolution time O~(s/ε)\widetilde{O}(s/\varepsilon)O(s/ε), matching the best known scaling while allowing a near-maximal step size Θ~(1/Λ)\widetildeΘ(1/Λ)Θ(1/Λ). As an independent contribution, we strengthen the evolution-time lower bound to Ωk(s1−1/(2k)/ε)Ω_k(s^{1-1/(2k)}/\varepsilon)Ωk​(s1−1/(2k)/ε), placing our algorithm within a factor O~(s1/(2k))\widetilde{O}(s^{1/(2k)})O(s1/(2k)) of being optimal. Eigenphase engineering departs from approaches based on dynamical approximation or iteratively canceling the unknown Hamiltonian. It encodes energy expectation differences in the perturbative response of eigenphases of the actual controlled evolution, then extracts them through phase estimation and high-order extrapolation. This avoids the need for polynomially shorter evolution steps as the target precision improves. A version using only single-qubit operations retains the same step size with a n\sqrt nn​ overhead in ℓ2\ell_2ℓ2​ learning time for nnn qubits. The method also enables tolerant Hamiltonian certification and estimation of a specified Pauli coefficient, the latter without locality or sparsity assumptions. These results establish eigenphase engineering as a versatile approach to extracting Hamiltonian information with infrequent control.