Efficient and robust entanglement generation with deep reinforcement learning for quantum metrology

Abstract

Quantum metrology exploits quantum resources and strategies to improve measurement precision of unknown parameters. One crucial issue is how to prepare a quantum entangled state suitable for high-precision measurement beyond the standard quantum limit. Here, we propose a scheme to find optimal pulse sequence to accelerate the one-axis twisting dynamics for entanglement generation with the aid of deep reinforcement learning (DRL). We consider the pulse train as a sequence of π/2 pulses along one axis or two orthogonal axes, and the operation is determined by maximizing the quantum Fisher information using DRL. Within a limited evolution time, the ultimate precision bounds of the prepared entangled states follow the Heisenberg-limited scalings. These states can also be used as the input states for Ramsey interferometry and the final measurement precisions still follow the Heisenberg-limited scalings. While the pulse train along only one axis is more simple and efficient, the scheme using pulse sequence along two orthogonal axes show better robustness against atom number deviation. Our protocol with DRL is efficient and easy to be implemented in state-of-the-art experiments.

Entanglement Generation of Polar Molecules via Deep Reinforcement Learning

“…Generally, RL is used to address sequential decision-making problems, in which an abstract agent learns an optimal decision strategy by iteratively interacting with the environment without any prior knowledge. 41 RL has already been widely applied to QIP over the past several years, including quantum state preparation, 42−45 quantum gate engineering, 46,47 quantum metrology, 48 quantum communication, 49,50 quantum heat engine, 51,52 quantum state transfer, 53,54 quantum state ansatz, 55 and quantum control. 56−58 At present, the deep RL technique combining with deep neural networks (DNNs) is actually more popular compared to the classic RL.…”

“…Generally, RL is used to address sequential decision-making problems, in which an abstract agent learns an optimal decision strategy by iteratively interacting with the environment without any prior knowledge . RL has already been widely applied to QIP over the past several years, including quantum state preparation, − quantum gate engineering, , quantum metrology, quantum communication, , quantum heat engine, , quantum state transfer, , quantum state ansatz, and quantum control. − At present, the deep RL technique combining with deep neural networks (DNNs) is actually more popular compared to the classic RL. It takes full advantages of the powerful representation of DNNs and can efficiently deal with challenging high-dimensional optimization problems. , In ref , the deep RL technique has been demonstrated to be superior to traditional optimal control methods for manipulating and controlling multilevel dissipative quantum systems.…”

Quantum metrology with superposition of GHZ state and Twin-Fock state

“…Robust and high-precision quantum metrology plays an important role in the research of quantum information science [1][2][3] and advanced quantum technologies, including gravitational-wave detection [4], atomic clocks [5][6][7], magnetometers [8][9][10], and biological measurement [11]. Tremendous efforts have been made to achieve this goal, ranging from the optimal choice of metrological useful quantum resources [12][13][14][15] to the meticulous design of measurement schemes [16][17][18][19][20][21]. Very nearly, a special superposition of GHZ state and Twin Fock (SGT) state has been realized in the experimental generation of superposed Dicke states [22].…”

Characterizing entanglement robustness of N-qubit W superposition state against particle loss from quantum Fisher information

“…Entanglement, as a key quantum resource, plays an indispensable role in quantum information science and quantum technology, including quantum metrology [1,2], quantum key distribution [3], quantum computation [4], quantum cryptography [5], and etc. Due to the fragile property of entanglement to the surrounding environment, many endeavors have been devoted to pursue a robust entangled state [6][7][8][9][10]. Meanwhile, the factor influencing the entanglement robustness also becomes an important issue [11][12][13].…”