Quantum control based on real-time environment analysis by spectator qubits
Abstract
The robust control of quantum systems is often limited by uncertainties in the applied controls and unwanted interactions with a dynamic environment. Macroscopic sensors lack sufficient correlation with the qubit environment to accurately assess these noise sources. The noise can be characterized by measuring the qubit directly, but the resulting calibration can become outdated as the noise changes over the course of a long experiment. A spectator qubit provides a probe that not only observes a noise environment similar to the data qubit but can also be probed during quantum operations on the data qubit. The spectator qubit allows us to update our model of the noise in real-time and provide feedback to the data qubit during the experiment. The goal of our team is to test the applicability of spectator qubits in real systems. We have gathered a team of experimentalists and theorists comprising of a US team (Berkeley, Duke, Johns Hopkins/APL, LSU, MIT, Oregon), an AUS team (Griffith, UNSW, UTS) and two unfunded collaborators (Dartmouth, SC Solutions). The experimental groups cover a wide range of qubit implementations: atomic ions (Haeffner, UC Berkeley), NV Centers (Wang, Oregon), superconductors (Oliver, MIT), and doped Si (Morello, UNSW). These groups are among the top academic groups in their respective domains. The theory groups are well-known in the area of quantum control with experts in dynamic decoupling(Dowling, LSU; Viola, Dartmouth), optimal control sequences (Brown, Duke; Wiseman, Griffith; Kosut, SCSolutions ), machine learning (Ferrie, UTS; Lucarelli, ffiU/APL), and noise characterization (Paz-Silva, Griffith; Quiroz and Schultz ffiU/APL). Together this group of theorists will develop new methods utilizing spectator qubits and theoretical limits on the use of spectator qubits performance. Our technical approach is based on overcoming three challenges associated with spectator qubits: 1) Experimental Integration, 2) Limited Bandwidth of Measurement, and 3) Optimizing Control Based on Statistical Information. Experimental integration requires that the spectator qubit be sufficiently close to the data qubit to sense the same environment, but not add too much additional noise...
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Feb 14, 2019
- Source ID
- W911NF1810218
Entities
People
- Kenneth R. Brown
Organizations
- Army Contracting Command
- Duke University
- United States Army