Techniques for noise suppression and robust control in spin-based quantum information processors

[摘要] Processing information quantum mechanically allows the relatively efficient solution of many important problems thought to be intractable on a classical computer. A primary challenge in experimentally implementing a quantum information processor is the control and suppression of environmental noise that decoheres the quantum system and causes it to behave classically. Environmental errors may be dynamically suppressed by applying coherent control pulses to the qubits that decouple the environment. However, the pulses themselves are subject to implementation errors, which hinders the ability to robustly store a complete quantum state. This thesis details results on the use of optimal control theory, noise twirling, and logical qubit encodings to design high-fidelity control pulses and decoupling sequences that are robust to implementation errors. Results are also presented that demonstrate how high-fidelity inductive control of a quantum system may be obtained with limited resonator bandwidth, with a discussion of applications to actuator-based quantum information processors. In a multi-mode design for such a processor, which allows efficient removal of entropy, a new protocol is suggested that permits robust parallel information transfer between nodes. The results detailed in this thesis apply broadly to most implementations of quantum information processing and specifically enable a new design for a spin-based multinode quantum information processor based on single-crystal molecular monolayer electron-nuclear spin systems integrated with superconducting electronics.

[发布机构] Massachusetts Institute of Technology