Singularity-free Quantum Control Enables High-Fidelity State Preparation in Non-Markovian Systems

State preparation represents a fundamental challenge in quantum technologies, crucial for advancements in computation and communication, and its difficulty intensifies in real-world systems affected by environmental noise. Ritik Sareen, Akram Youssry, and Alberto Peruzzo, from RMIT University and Quandela, now present a new method for precisely controlling quantum states even under complex noise conditions. Their approach overcomes limitations of existing techniques by generating control pulses that avoid problematic singularities, ensuring experimental feasibility, and works effectively whether the noise is fully understood or remains largely unknown. The team demonstrates that this framework achieves high-fidelity state preparation across a range of quantum targets, offering a versatile pathway toward robust quantum control on current and future quantum devices exhibiting complex, non-Markovian behaviour.

Optimizing Quantum Control Under Noise

This research details a project focused on optimizing quantum control pulses to reliably manipulate qubits, even when subjected to environmental disturbances. Scientists are comparing different strategies for achieving high-fidelity state control, considering scenarios where the noise characteristics are known or completely unknown. Real-world quantum systems inevitably experience noise, which degrades performance, so developing robust control strategies is essential. The work investigates how to best design control pulses to achieve accurate and predictable quantum state transformations despite the presence of noise.

Visualizing qubit state evolution using the Bloch sphere helps track how control pulses manipulate the quantum state. The team demonstrates that a graybox optimization strategy, designed for unknown noise, is crucial for achieving high fidelity in realistic quantum systems. By comparing optimized pulses to those representing worst-case and average-case scenarios, scientists establish a benchmark for assessing their effectiveness. The optimization results vary depending on the desired final state, highlighting the need for adaptable control strategies. In summary, this research presents a comprehensive study of quantum control optimization, emphasizing robustness to noise. Scientists combine theoretical analysis, numerical simulations, and experimental data to develop and evaluate different optimization algorithms, paving the way for more reliable quantum technologies.

Robust Single-Qubit Control Against Realistic Noise

Scientists have developed a new protocol for preparing quantum states in single qubits with high accuracy, even when those qubits are subject to realistic noise. This work addresses a significant challenge in quantum computing, where maintaining the delicate quantum state of a qubit is hampered by environmental disturbances. The team constructs a family of physically realizable control pulses, guaranteeing finite amplitude. To extend control to real-world quantum systems, an optimal control layer selects the most robust pulse, minimizing the impact of noise. Numerical simulations demonstrate the protocol’s effectiveness under multi-axis classical colored noise, achieving high-fidelity state preparation for arbitrary qubit targets.

This method differs from gradient-based optimization algorithms by directly constructing admissible pulses through algebraic relations, avoiding iterative searches and potential convergence to local minima. The team rigorously proved that all pulses generated by their method are bounded, avoiding singular control pulses that limit the physical feasibility of many existing approaches. The framework naturally incorporates experimental constraints like finite bandwidth, and the embedded optimal control and machine-learning modules provide adaptability to complex, partially understood noise environments.

Bounded Control Pulses for Robust Qubit Preparation

Scientists have developed a new protocol for preparing quantum states in qubits, even when those qubits are subject to complex, realistic noise. This work addresses a critical challenge in quantum computing, where maintaining the delicate quantum state of a qubit is hampered by environmental disturbances. The team’s approach guarantees the creation of control pulses that remain finite in amplitude, avoiding physically unrealistic, singular pulses often produced by other methods. The research centers on a two-stage process. First, scientists generate a family of control pulses capable of perfectly steering a closed quantum system from any initial state to a desired final state.

Crucially, these pulses are all mathematically guaranteed to be bounded, ensuring they can be physically implemented. Second, the team identifies the optimal pulse from this family, minimizing the impact of noise on the system’s evolution. This framework accommodates both characterized and uncharacterized noise, offering robustness even when the precise nature of the disturbance is unknown. Experiments demonstrate high-fidelity state preparation for arbitrary qubit targets under multi-axis classical colored noise. The team rigorously proved that all generated pulses are bounded, remaining finite in amplitude at all times.

The invariant-based method directly constructs families of admissible pulses through algebraic relations, guaranteeing physical realizability and bounded amplitudes. The framework naturally incorporates constraints such as finite bandwidth and experimental limitations. This work represents an important step toward realizing high-performance quantum operations on near-term quantum devices.

Robust Quantum Control Against Environmental Noise

This research presents a new method for precisely controlling quantum systems, even when those systems are affected by environmental noise. Scientists developed a two-stage control process that first creates a set of possible control pulses capable of achieving a desired quantum state in an ideal scenario. The method then identifies the pulse within this set that is most resilient to noise, minimizing errors in the final quantum state. This approach successfully prepares quantum states with high accuracy, while generating control signals that are practical for implementation on current quantum computing hardware.

The framework extends existing control techniques by accommodating both known and unknown sources of noise, offering a versatile tool for robust quantum state engineering. Importantly, the team demonstrated that the generated control pulses remain physically realistic, avoiding the singularities or infinite values that can plague other methods. They achieved this by carefully bounding the magnitude of the control parameters and ensuring the continuity of key functions throughout the control process.