Noise-aware Time-optimal Quantum Control

Abstract: Quantum optimal control plays a vital role in many quantum technologies, including quantum computation. One of the most important control parameters to optimise for is the evolution time (pulse duration). However, most existing works focus on finding the shortest evolution time theoretically possible without offering explicit pulse constructions under practical constraints like noise in the system. This paper addresses these limitations by introducing an efficient method to perform the Chopped Random Basis (CRAB) optimisation in the presence of noise, specifically when the noise commutes with the gate Hamiltonian. This noise-aware approach allows for direct optimisation of the evolution time alongside other control parameters, significantly reducing the computational cost compared to full noisy simulations. The protocol is demonstrated through numerical simulations on state-to-state transfer and gate compilation problems under several noise models. Results show that the optimised fidelity has a strong dependence on evolution time due to noise, drift Hamiltonian, and local traps in optimisation, highlighting the necessity of optimising evolution time in practical settings that can lead to a substantial gain in the fidelity. Our pulse optimisation protocol can consistently reach the global optimal time and fidelity in all of our examples. We hope that our protocol can be the start of many more works on the crucial topic of control pulse time optimisation in practical settings.