arXiv:2610.09666
Geodesic-based optimal control for leakage suppression in superconducting qubits
Adonai Hilário da Silva, Tanmoy Pandit, Francesco Cosco
Abstract
Superconducting qubits are often based on the first two energy levels of a physical anharmonic LC oscillator. Since the anharmonicity of such systems is usually much lower than the involved energy gaps, population can easily leak to higher energy states. This is what characterizes leakage errors, and it can have multiple sources: from coherent error induced by the control platform to incoherent thermal dissipation. In this work, we adapt an optimal control method based on sub-Riemannian geodesic search in the unitary group for a system consisting of two superconducting qubits with leakage levels. We use the $R_x(π/2)$ rotation for single-qubit operation analysis, and the $i$SWAP operation for two-qubit analysis. For the former, we find considerable performance improvement in average gate fidelity as well as in the reduction of leakage errors when comparing the obtained optimal pulses with Gaussian and DRAG pulses, especially for short duration pulses. For the $i$SWAP gate, although leakage error remains roughly the same, we manage to obtain considerably higher fidelities with the cost of requiring a slightly longer time evolution.
AI Summary
The paper investigates whether sub-Riemannian geodesic optimal control can suppress leakage while implementing gates in superconducting transmons. It adapts an existing energy-minimizing geodesic-search method by multiplying the control Hamiltonian by a smooth envelope that brings the pulses close to zero at their endpoints. Pulse synthesis uses numerical shooting over initial costates, followed by process-fidelity optimization. The model comprises two Duffing-oscillator qubits coupled directly and through a tunable coupler. A dispersive Schrieffer–Wolff transformation eliminates the coupler, and three-level truncation gives an effective $\mathrm{SU}(9)$ control problem. Performance is evaluated numerically with thermal relaxation and excitation, using average gate fidelity, leakage, and seepage. For $R_x(\pi/2)$, the optimized pulses outperform Gaussian pulses and two fixed-parameter DRAG variants at durations below approximately $20$ ns. A $17$ ns pulse achieves reported average fidelity $0.9996$, average leakage $1.08\times10^{-5}$, and seepage $2.12\times10^{-3}$, reaching the simulated thermal-noise floor. For $i$SWAP, microwave controls increase reported fidelity from $0.9548$ at $79.5$ ns without controls to $0.9942$ at $86$ ns. Leakage increases slightly, from $4.55\times10^{-4}$ to $6.69\times10^{-4}$, while seepage remains nearly unchanged. The results are simulations rather than hardware demonstrations. Practical implementation is stated to depend on waveform-generation precision and Hamiltonian-parameter calibration. Optimization retains only the first leakage level; subsequent checks including higher levels find negligible changes. Connections to higher-order DRAG corrections and extension to flux-control protocols remain open.
Central claim may not hold: The stated process-fidelity formula gives average fidelity 13 for a perfect two-qubit identity channel, making the reported fidelity benchmarks unverifiable without clarification of the implemented formula. Multiplying the geodesic control by the envelope S(t) does not establish optimality under the paper's stated unweighted energy cost and instead admits a different, time-weighted variational interpretation. The supplied parameters and coupling formulas give residual same-transition couplings near 38 kHz at the stated decoupling frequency, rather than the approximately 240 Hz claimed.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 16 pages, 6 figures