Spectral Analysis and Identification of Noises in Quantum Systems

• Published in: arXiv.org 2012-11
• Main Authors: Re-Bing Wu, Tie-Fu, Li, Kofman, A G, Zhang, Jing, Yu-xi, Liu, Pashkin, Yu A, Jaw-Shen Tsai, Nori, Franco
• Format: Article
• Language: English
• Subjects: Approximation; Data processing; Markov processes; Mathematical analysis; Noise; Noise measurement; Quantum phenomena; Quantum theory; Qubits (quantum computing); Superconductivity
• ISSN: 2331-8422
• DOI: 10.48550/arxiv.1211.5186

In quantum information processing, knowledge of the noise in the system is crucial for high-precision manipulation and tomography of coherent quantum operations. Existing strategies for identifying this noise require the use of additional quantum devices or control pulses. We present a noise-identification method directly based on the system's non-Markovian response of an ensemble measurement to the noise. The noise spectrum is identified by reversing the response relationship in the frequency domain. For illustration, the method is applied to superconducting charge qubits, but it is equally applicable to any type of qubits. We find that the identification strategy recovers the well-known Fermi's golden rule under the lowest-order perturbation approximation, which corresponds to the Markovian limit when the measurement time is much longer than the noise correlation time. Beyond such approximation, it is possible to further improve the precision at the so-called optimal point by incorporating the transient response data in the non-Markovian regime. This method is verified with experimental data from coherent oscillations in a superconducting charge qubit.