Feedback-based Steering for Quantum State Preparation

Volya, Daniel, Pan, Zhixin, Mishra, Prabhat

2023 IEEE International Conference on Quantum Computing and Engineering (QCE), pages 1308–1318, September 2023, doi: 10.1109/QCE57702.2023.00148

pdf

Abstract

State preparation is an essential component in quantum information science. A recently developed steering protocol utilizes a sequence of generalized measurements on a detector to steer a quantum system towards a desired state. However, it is designed as an open-loop technique that requires accurate modeling of the overall quantum system and can be prone to errors. To address this challenge, we propose a closed-loop control technique that introduces feedback to the steering protocol, providing robustness to noise and faster state convergence. We introduce two strategies for feedback: (1) a gradient-based active steering protocol that changes the detector-system coupling conditioned on the detector’s readout and (2) tuning the fixed detector-system coupling via model-free reinforcement learning. We study the effectiveness of these strategies under various noise models, including both incoherent and decoherent noise, and discuss potential applications in quantum technologies.

Bibtex


@inproceedings{volyaFeedbackQuantumSteering2023,
  title = {Feedback-based Steering for Quantum State Preparation},
  booktitle = {2023 IEEE International Conference on Quantum Computing and Engineering (QCE)},
  author = {Volya, Daniel and Pan, Zhixin and Mishra, Prabhat},
  year = {2023},
  month = {sept},
  pages = {1308--1318},
  doi = {10.1109/QCE57702.2023.00148},
  abstract = {State preparation is an essential component in
  quantum information science. A recently developed steering
  protocol utilizes a sequence of generalized measurements on a
  detector to steer a quantum system towards a desired state.
  However, it is designed as an open-loop technique that requires
  accurate modeling of the overall quantum system and can
  be prone to errors. To address this challenge, we propose
  a closed-loop control technique that introduces feedback to
  the steering protocol, providing robustness to noise and faster
  state convergence. We introduce two strategies for feedback:
  (1) a gradient-based active steering protocol that changes
  the detector-system coupling conditioned on the detector’s
  readout and (2) tuning the fixed detector-system coupling via
  model-free reinforcement learning. We study the effectiveness
  of these strategies under various noise models, including
  both incoherent and decoherent noise, and discuss potential
  applications in quantum technologies.},
  keywords = {Quantum computing, quantum measurement,
  quantum steering, state preparation, quantum control},
}