Skip to content
Quantum.world

The Signal

Researchers publish simulated feedback protocols for quantum drift

A paper published in Physical Review Applied describes two feedback protocols for adjusting quantum-computer controls as they drift. The authors tested them in numerical simulations, including a five-qubit error-correction code using syndrome data.

Researchers publish simulated feedback protocols for quantum drift

Two routes to in-operation calibration

Two simulated feedback methods for correcting drifting quantum-computer controls are now described in a paper published in Physical Review Applied on November 18. The authors report that one method can adjust a parameter after each measurement, while a second uses outcomes from circuits expected to produce a definite result.

The paper’s authors are Alicia B. Magann, Nathan E. Miller, Robin Blume-Kohout, Peter Maunz and Kevin C. Young. Their work sets out protocols for updating control parameters during operation rather than relying only on collecting a batch of measurements before making a correction.

Updates after measurement

The first method uses what the paper calls indefinite-outcome circuits: when the circuit and controls are ideal, measurement results are distributed uniformly across possible outcomes. A departure from that distribution signals a control error, and the protocol changes the parameter after each individual measurement. In a single-qubit example, the ideal circuit gives either result with equal, 50% probability.

The second method uses definite-outcome circuits, which should return a particular result when implemented ideally. The protocol gathers measurements until it observes a set number of unexpected outcomes, then uses them to estimate the size of a parameter error. Because that estimate does not reveal whether the parameter should move up or down, the method alternates its assumed correction direction; if a change worsens the error, a later adjustment can reverse it.

The authors say the two approaches are intended for different kinds of circuits. Their paper gives syndrome-extraction circuits used in quantum error correction as an example for the definite-outcome method, because the syndrome measurements can provide feedback during error-correction operation.

Simulations, not a hardware trial

Numerical simulations tested calibration of one- and two-qubit gates in the presence of parameter drift, decoherence and state-preparation and measurement errors. The researchers also simulated feedback during quantum error correction using syndrome data from the [[5,1,3]] code, a five-qubit code. The paper describes that demonstration as evidence of feasibility in simulation, not a result from a quantum processor.

For that error-correction simulation, the paper specifies an idealized code-capacity model: it excludes stochastic errors such as decoherence and readout errors, and assumes syndrome extraction itself is flawless. The authors say further work would be needed to move toward a practical laboratory method, including modeling noisy operations and erroneous ancilla readout.

Fast updates also depend on the classical control hardware. The paper notes that some systems cannot change control parameters after each experimental run, while certain FPGA-based controllers can adapt waveforms almost instantly. It says systems with slower updates would need to balance data collection against the time required to apply corrections.

The article was first posted to arXiv on December 8, 2025, and later appeared in the journal. Its results remain based on numerical simulations.

Sources

  1. Physical Review Applied
  2. arXiv

Related stories

Explore

More articles