Quantum optimal control robust to $1/f^α$ noises using fractional calculus: voltage-controlled exchange in semiconductor spin qubits

Abstract: Low-frequency $1/f^\alpha$ charge noise significantly hinders the performance of voltage-controlled spin qubits in quantum dots. Here, we utilize fractional calculus to design voltage control pulses yielding the highest average fidelities for noisy quantum gate operations. We focus specifically on the exponential voltage control of the exchange interaction generating two-spin $\mathrm{SWAP}^k$ gates. When stationary charge noise is the dominant source of gate infidelity, we derive that the optimal exchange pulse is long and weak, with the broad shape of the symmetric beta distribution function with parameter $1-\alpha/2$. The common practice of making exchange pulses fast and high-amplitude still remains beneficial in the case of strongly nonstationary noise dynamics, modeled as fractional Brownian motion. The proposed methods are applicable to the characterization and optimization of quantum gate operations in various voltage-controlled qubit architectures.