Scientists have developed a universal protocol capable of designing arbitrary high-dimensional quantum states

The use of high-dimensional quantum states in the quantum information protocol can achieve better performance in applications ranging from secure quantum communication to fault-tolerant quantum computing. The development of a universal protocol capable of designing arbitrary high-dimensional quantum states will be a major achievement.
 
To this end, researchers have proposed and developed several strategies and platforms. Quantum walking (QW) dynamics has been shown to allow the development of universal, platform-independent state engineering protocols. However, the existence of unavoidable noise and defects in the characterization of experimental instruments reduce the overall quality of state generation.
 
In order to overcome these limitations, a group of researchers from the University of Rome I, Queen’s University of Belfast and the University of Palermo demonstrated an adaptive optimization protocol that can design arbitrary high-dimensional quantum states. The research results have been published in Advanced Photonics journal. In the case of a complete black box, the protocol only relies on the measurement consistency between the generation state and the target state to adjust the relevant experimental parameters, without the need to describe the generation settings.

  
Design an automated platform scheme for arbitrary quantum states in the angular momentum of the photon orbit. The experimental parameters are      adaptively optimized to generate the target state

The author uses classical light and single-photon orbital angular momentum (OAM) to experimentally verify the proposed protocol. OAM is the degree of freedom of the electromagnetic field related to its space and phase profile. Since OAM is an infinite degree of freedom, it is suitable for encoding arbitrary high-dimensional quantum states. The author experimentally implemented the protocol using a state generation platform based on quantum walk dynamics and polarization degrees of freedom. By adjusting the parameters of the operator acting on the polarization state, it is possible to design any walking state encoded in the OAM space. Then, the proposed optimization algorithm adjusts the experimental parameters of driving dynamics online to obtain the desired results.
                 

laboratory apparatus. (a) The engineering protocol has been experimentally tested in the three-step discrete time QW coded in the optical OAM, including single-photon input and a classic continuous wave laser (CNI laser PSU-III-FDA) with a wavelength of 808 nm. The single-photon state is produced by the type II spontaneous parametric down-conversion process in the periodically polarized KTP crystal. The input state is characterized by horizontal polarization and OAM characteristic value m=0. Each step of QW is completed by a coin operator, implemented by a set of wave plates (QWP-HWP-QWP), and the shift operator is implemented by QP. (b) During the algorithm iteration, RBFOpt samples the black box function to construct an alternative model for optimization. In the kth iteration, the algorithm receives the fidelity calculated in the previous iteration as input and uses it to improve the agent modeling. In addition, a new parameter θk is calculated based on the optimization process. This process is repeated for    each iteration of the algorithm.

 
For several four-dimensional target OAM states, the optimization protocol performed well under experimental noise conditions. Finally, the research team studied the adaptability of the protocol by introducing time-varying noise as the external disturbance of the parameter value. After introducing these external disturbances, the agreement found a new optimal solution. The proposed protocol is suitable for many situations, even if there is interference, no major fine-tuning is required.
 

Simulation optimization: 1-F of fidelity obtained in different stages of optimization. The algorithm is tested on 10 random target states, and the optimization is repeated 10 times for each state. The reported result is the average behavior of each of the 10 states. The highest average fidelity          obtained is 0.994±0.002. The shaded area represents the standard deviation of the mean

According to senior author Fabio Sciarrino, head of the Quantum Information Laboratory of the Department of Physics in Rome, “The proposed dynamic learning protocol will help to complete some quantum information tasks that require finding the optimal experimental parameters under noisy conditions. value."