Parameterized Quantum Circuits for Learning Cooperative Quantum Teleportation

Abstract

Quantum teleportation is an elemental process in quantum communication and many variants have been widely investigated theoretically and experimentally. Motivated by cooperation in classical communications, cooperative quantum teleportation (CQT) are developed with features of the cooperation referring to allocation of resources and operations among participants. Parameterized quantum circuit (PQC) are employed to learn the CQT protocols on account of different training scenarios with controls of gate parameters among shared entangled states, measurement, and recovery operations. Numerical results show the CQT protocols can be discovered with unit fidelity regardless of ideal or nonideal channels by training the PQCs with hybrid quantum‐classical (HQC) optimization under full cooperations of the participants. Further, the influence of quantum channel noise on the teleportation performance is explored by parameterizing the entangled states with noise and entanglement parameters under adjustable recovery operations. Simulation results indicate the trained PQC can improve the system performance, which implies the potential denoising capability of the HQC algorithms. The suggested CQT protocols satisfy the underlying properties of universality, randomness, locality and cooperation, and discussions indicate the designed quantum circuits can be optimized in view of quantum and circuit costs by appropriate optimization algorithms for achieving higher state fidelity in the future work.

Dynamic Variational Quantum Eigensolver: Incorporating Time-evolution for Ansatz Design and Ground State Simulation

“…This research provides a valuable reference for the development of variational quantum algorithms. Furthermore, we provide effective ideas for designing parameterized quantum circuits [48,49].…”

Prediction of Protein‐Ligand Binding Affinity by a Hybrid Quantum‐Classical Deep Learning Algorithm

“…on hybrid quantum-classical algorithms, [19][20][21][22][23][24][25] which can be realized on classical computers by constructing a suitable software framework. [26] To date, many hybrid quantum-classical machine learning algorithms have been proposed to tackle problems in physics, [27] chemistry, and engineering, [28,29] but their application on prediction of protein-ligand binding affinity has been reported only rarely. In this work, we have attempted to introduce quantum algorithms into classical deep learning.…”

“…To date, many hybrid quantum‐classical machine learning algorithms have been proposed to tackle problems in physics, [ 27 ] chemistry, and engineering, [ 28,29 ] but their application on prediction of protein‐ligand binding affinity has been reported only rarely. In this work, we have attempted to introduce quantum algorithms into classical deep learning.…”

Chaotic Image Encryption Based on Boson Sampling

“…Fortunately, with the rapid development of quantum information theory, [ 10–13 ] researches have proved that the quantum system also has chaotic characteristics, [ 14,15 ] which provides a new idea for solving the problems existing in some classical chaotic systems. Akhshani et al.…”

“…[8] Unfortunately, most chaotic sequences generated by classical chaotic systems are unstable due to the periodicity of chaotic mapping, causing image encryption methods based on these principles are no longer indestructible. [9] Fortunately, with the rapid development of quantum information theory, [10][11][12][13] researches have proved that the quantum system also has chaotic characteristics, [14,15] which provides a new idea for solving the problems existing in some classical chaotic systems. Akhshani et al proposed a new image encryption algorithm applying quantum logistic map to the construction of a chaotic cryptographic system.…”