High-Resolution Multilevel Noise Spectroscopy via Fixed-Time Control

In superconducting qubit architectures, characterizing noise beyond the qubit manifold is essential for diagnosing leakage mechanisms and enabling high-fidelity multilevel gate operations. When noise exhibits spatiotemporal correlations, quantum noise spectroscopy (QNS) provides a powerful means for characterizing its spectral properties. Extensions of QNS protocols to n-level systems have shown promise, some of which have been experimentally implemented. However, existing approaches—often based on continuous driving or comb-based modulation—are optimized for smooth spectra and struggle to resolve sharp or structured noise features without substantially increasing experimental overhead. In this work, we extend fixed-total-time pulse sequences (FTTPS), previously developed for qubit QNS, to the qutrit regime. Through numerical simulations, we identify parameter regimes in which FTTPS offers superior spectral sensitivity compared to comb-based protocols. Furthermore, we demonstrate that integrating classical signal processing techniques enhances spectrum reconstruction accuracy. Together, these results establish a new multilevel noise characterization protocol that improves spectral estimation while remaining compatible with current superconducting qubit hardware.

LLNL-ABS-2012794