Abstract:It is well-known that simulating quantum circuits with low but non-zero hardware noise is more difficult than without noise. It requires either to perform density matrix simulations (coming with a space overhead) or to sample over "quantum trajectories" where Kraus operators are inserted randomly (coming with a runtime overhead). We propose a simulation technique based on a representation of hardware noise in terms of trajectories generated by operators that remain close to identity at low noise. This representation significantly reduces the variance over the quantum trajectories, speeding up noisy simulations by factors around $10$ to $100$. As a by-product, we provide a formula to factorize multiple-Pauli channels into a concatenation of single Pauli channels.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to improve efficiency in simulating noisy quantum circuits. Specifically, the author proposes a new simulation technique. By systematically inserting small perturbations after each gate instead of occasionally inserting large perturbations to generate trajectories, the variance between different trajectories is significantly reduced, and the simulation of noisy quantum circuits is accelerated. This method can provide a running - time acceleration of 1 - 2 orders of magnitude while keeping the class of classically simulatable circuits unchanged. ### Background and Problem of the Paper As the capabilities of quantum computers steadily increase, both the number of qubits and the gate fidelity are constantly improving. However, for application - oriented quantum engineers, dealing with hardware noise remains a long - standing challenge. Besides accurately simulating a small number of qubits, noisy simulation is crucial for optimizing the noise - mitigation part in algorithm implementation. However, it is well - known that simulating quantum circuits with low but non - zero noise is more difficult than noiseless simulation. This usually requires performing density - matrix simulation (resulting in space overhead) or sampling "quantum trajectories" where Kraus operators are randomly inserted (resulting in running - time overhead). ### Proposed Method The paper proposes a new simulation technique based on trajectory generation. These trajectories are generated by operators close to the identity operator, thus significantly reducing the variance between trajectories and speeding up the noisy simulation. Specifically, the author proposes a new method. Instead of inserting an $X$ error with probability $\epsilon$, a small random rotation $e^{i\theta X}$ is systematically inserted after each gate. This new method can significantly reduce the variance of the expected value, thereby reducing the number of trajectories that need to be simulated. ### Technical Details 1. **Digital Sampling vs. Analog Sampling**: - Digital Sampling: Kraus operator insertion is usually used, such as $W = \begin{cases}I, & \text{probability } 1-\epsilon\\\bar{K}_q, & \text{probability }\epsilon p_q / \sum_{q' = 1}^Q p_{q'}\end{cases}$. - Analog Sampling: A small random rotation $e^{i\theta X}$ is systematically inserted, where the variance of $\theta$ changes with $\epsilon$. 2. **Factorization of Multiple - Pauli - String Channels**: - The paper also provides a formula to factorize any multiple - Pauli - string channel into the concatenation of multiple single - Pauli - string channels. 3. **Numerical Implementation**: - The author demonstrates the effectiveness of analog sampling through numerical experiments. In particular, when calculating the expected value of local observables and sampling from the wave function, analog sampling provides a 10 - to - 100 - fold acceleration compared to digital sampling. ### Conclusion The paper proposes an efficient noisy - quantum - circuit - simulation technique. By systematically inserting small perturbations to generate trajectories, the simulation efficiency is significantly improved. This method is not only innovative in theory but also shows significant advantages in practical applications.

Analog simulation of noisy quantum circuits

Pauli path simulations of noisy quantum circuits beyond average case

Simulating Quantum Pauli Noise with Three Independently Controlled Pauli Gates

Classical simulations of noisy variational quantum circuits

Simulating photonic devices with noisy optical elements

Simulating Noisy Quantum Circuits with Matrix Product Density Operators

Simulating Noisy Variational Quantum Algorithms: A Polynomial Approach

Approximation Algorithm for Noisy Quantum Circuit Simulation

Modeling and simulation of a quantum thermal noise on the qubit

Simulating Quantum Mean Values in Noisy Variational Quantum Algorithms: A Polynomial-Scale Approach

Noise-induced shallow circuits and absence of barren plateaus

Efficient Classical Simulation of Noisy Quantum Computation

A polynomial-time classical algorithm for noisy quantum circuits

Improved Digital Quantum Simulation by Non-Unitary Channels

Logical Abstractions for Noisy Variational Quantum Algorithm Simulation

Modeling and simulating the noisy behavior of near-term quantum computers

Effect of non-unital noise on random circuit sampling

Efficient classical simulation and benchmarking of quantum processes in the Weyl basis

A novel approach to noisy gates for simulating quantum computers

Classically simulating quantum circuits with local depolarizing noise