부산대학교 도서관

In quantum computing, characterising the full noise profile of qubits can aid the efforts towards increasing coherence times and fidelities by creating error mitigating techniques specific to the type of noise in the system, or by completely removing the sources of noise. Spin qubits in MOS quantum dots are exposed to noise originated from the complex glassy behaviour of two-level fluctuators, leading to non-trivial correlations between qubit properties both in space and time. With recent engineering progress, large amounts of data are being collected in typical spin qubit device experiments, and it is beneficiary to explore data analysis options inspired from fields of research that are experienced in managing large data sets, examples include astrophysics, finance and climate science. Here, we propose and demonstrate wavelet-based analysis techniques to decompose signals into both frequency and time components to gain a deeper insight into the sources of noise in our systems. We apply the analysis to a long feedback experiment performed on a state-of-the-art two-qubit system in a pair of SiMOS quantum dots. The observed correlations serve to identify common microscopic causes of noise, as well as to elucidate pathways for multi-qubit operation with a more scalable feedback system.
Comment: updated reference