Quantum Control Reveals Hidden Noise Patterns

Researchers at the Centre for Quantum Computation and Communication Technology have developed a new control-centric approach to quantum noise spectroscopy, focusing on time-ordered polyspectra, which significantly improves methods for characterising environmental noise impacting quantum systems. Kaiah Steven and colleagues present a set of tools that removes constraints on control filter functions and enables broader applicability to various control scenarios. Simulations successfully reconstruct these polyspectra in both classical and quantum environments, representing a key step towards high-fidelity quantum control and targeted decoherence suppression for computing and sensing applications.

Reduced error in polyspectra reconstruction unlocks detailed quantum environment analysis

Error rates in reconstructing time-ordered polyspectra dropped by a factor of five compared to existing methods, enabling accurate characterisation of quantum environments previously beyond reach. This improvement stems from a new control-centric approach to quantum noise spectroscopy (QNS), which recasts the problem to focus on time-ordered polyspectra and eliminate artificial constraints on control filter functions. These functions are crucial as they shape the control signals used to probe the environment; by removing these limitations, generalised frequency-comb QNS protocols function effectively across a broader range of control scenarios. This is akin to correcting a warped reflection, allowing for a more accurate representation of the underlying noise. Traditional QNS methods often impose constraints on these filter functions to simplify the mathematical analysis, but these constraints can introduce symmetries into the reconstructed noise spectrum, distorting the true environmental characteristics. The new approach avoids this by directly optimising the control signals for noise estimation, rather than imposing pre-defined shapes.

Simulations successfully reconstructed these polyspectra across both classical Gaussian and quantum non-Gaussian environments, validating the framework’s ability to characterise complex noise. The ability to accurately reconstruct polyspectra in both scenarios demonstrates the versatility of the method. Classical Gaussian noise represents a common approximation for many environmental disturbances, while quantum non-Gaussian noise arises from more exotic sources and is particularly challenging to characterise. The method naturally accounts for non-commuting dynamics arising from real-world experimental limitations such as finite pulse widths and multi-axis interactions. Finite pulse widths, inherent in any physical implementation, introduce uncertainties in the timing of control operations, while multi-axis interactions, where control signals affect multiple degrees of freedom simultaneously, complicate the relationship between the control signals and the system’s response. By explicitly modelling these effects, the new approach provides a more realistic and accurate characterisation of the quantum environment. Precise characterisation of environmental noise is vital for realising practical quantum technologies, enabling mitigation of decoherence and improvement of control, and this is now more achievable than ever. Decoherence, the loss of quantum information due to interaction with the environment, is a major obstacle to building stable quantum computers and sensors; understanding and mitigating this noise is therefore paramount.

Non-parametric quantum noise spectroscopy techniques currently offer a model-agnostic approach to understanding these disruptive influences, but refocusing the analysis on control parameters improves noise estimation in challenging conditions. Unlike parametric methods, which assume a specific model for the noise, non-parametric techniques estimate the noise spectrum directly from the data, without making any prior assumptions. This makes them more versatile and applicable to a wider range of environments. Discrepancies between the real and imaginary components of the polyspectra offer a diagnostic tool for identifying areas needing higher sampling resolution, and further enhancement of the technique could be achieved by exploring different pulse sequence designs. Optimising the pulse sequence design, the specific sequence of control signals applied to the quantum system, could further improve the accuracy and efficiency of the noise estimation process. The real and imaginary parts of the polyspectra contain complementary information about the noise; discrepancies between them can indicate regions where the noise is poorly characterised and requires more detailed sampling. Despite the fact that current simulations do not yet demonstrate performance with actual quantum hardware, and significant engineering challenges remain before practical implementation, the framework offers a pathway to broader application of existing noise spectroscopy techniques even when full control isn’t achievable. Implementing this on real hardware requires careful consideration of factors such as signal-to-noise ratio, control fidelity, and the complexity of the quantum system.

A key shift in how environmental noise impacting quantum systems is analysed is represented by adopting a control-centric viewpoint. This allows for a more direct connection between the control signals and the estimated noise spectrum, leading to improved accuracy and efficiency. Traditionally focused on estimating spectral properties, such as the power spectral density of the noise, this approach reframes quantum noise spectroscopy by prioritising the characteristics of the control signals used to probe the environment. The method utilises M-fold periodic repetition of a pulse sequence with cycle time τ, allowing for broader application of existing techniques and improved estimation in complex scenarios; detailed records of noise events unfolding over time become central to the process, achieving this. By focusing on the time-ordered polyspectra, the method captures the correlations between different noise events, providing a more complete and accurate characterisation of the quantum environment. The periodic repetition of the pulse sequence allows for averaging over multiple cycles, reducing the impact of random noise and improving the signal-to-noise ratio. The cycle time τ determines the maximum frequency of noise that can be accurately characterised; shorter cycle times allow for the resolution of higher-frequency noise components. The polyspectra represent a higher-order statistical measure of the noise, capturing information about the joint probability distribution of multiple noise variables, which is often lost in simpler measures like the power spectrum.

The research demonstrated a new approach to analysing environmental noise affecting quantum systems by prioritising the characteristics of the control signals used to probe the environment. This control-centric viewpoint improves the accuracy of quantum noise spectroscopy and allows existing techniques to be applied even when complete control is not possible. By focusing on time-ordered polyspectra, the method captures correlations between noise events, providing a more complete characterisation of the quantum environment. The authors performed simulations across both classical and quantum environments to validate this framework.