Simulation of Open Quantum Systems via Low-Depth Convex Unitary Evolutions

Current methods for simulating open quantum systems either rely on classical calculations, which suffer from exponentially scaling runtimes, or quantum approaches that are typically impractical for near-term purposes due to the additional qubit and gate sequences needed and the associated noise. We propose a method that addresses this issue by decomposing a quantum noise channel into unitary operations and then applying these on a quantum computer. The method does not require deep ancilla frameworks and thus can be implemented with lower noise costs. The scheme can be efficiently run in parallel, and for random-unitary noise channels, this mapping is exact. For general noise channels, we present methods which approximately capture the non-unitary dynamics, allowing us to efficiently characterize the system evolution for a variety of systems.