Multiobjective Optimization for Robust Holonomic Quantum Gates

Abstract: Robust pulses have been widely used to reduce the sensitivity of quantum gate operations against various systematic errors due to the imperfections in practical quantum control. Yet, the typical optimization focuses on minimizing one type of errors serving as the one-objective algorithm, which arises a more susceptible sensitivity to other error sources. Optimizing multiple conflicting objectives of errors simultaneously remains a big challenge in quantum computing. Here, we propose a multiobjective optimization algorithm to achieve nonadiabatic holonomic quantum gates with enhanced robustness. We show that by considering the amplitude error, the detuning error and the decoherence of the Rydberg state as three individual objectives to be minimized, this algorithm can effectively balance multiple competing objectives, giving rise to a set of Pareto optimal solutions. We apply the Entropy Weight method to select the best solution that implements the robust holonomic gates, outperforming existing optimal gates with one-objective by having both higher gate fidelity and stronger robustness. This numerical approach of optimizing gates with multiple objectives can be readily applied to other gate protocols featuring a promising advance in fault-tolerant quantum computing with Rydberg atoms.