Dynamical Decoupling Sequences for Dealing With Time-Dependent Noise Spectrums

Techniques aiming to decouple a qubit from its noisy environment by applying particular control sequences are known as Dynamical Decoupling (DD). Many DD sequences have been proposed in the last decades, each with different control requirements and resulting in different filter functions of the environmental noise. One characteristic that all such sequences have in common, though, is the assumption of a noise spectrum of the environment that is stationary in time. This is a very reasonable hypothesis for spin ensembles or few-qubit systems. However, generic quantum algorithms will likely require periodic driving of many qubits that do not reflect the structure of a simple memory preservation experiment. Due to both DC and microwave crosstalk, control signals sent to the processor have a non-negligible effect on the noise spectrum felt by the qubits. In this talk, we show that tailoring a DD sequence where we choose not only the microwave pulse spacing but also the pulse positions within one algorithm cycle, reduces qubit dephasing. We perform numerical simulations and compare them to experimental data.