Abstract:Compared with time independent Hamiltonians, the dynamics of generic quantum Hamiltonians $H(t)$ are complicated by the presence of time ordering in the evolution operator. In the context of digital quantum simulation, this difficulty prevents a direct adaptation of time independent simulation algorithms for time dependent simulation. However, there exists a framework within the theory of dynamical systems which eliminates time ordering by adding a "clock" degree of freedom. In this work, we provide a computational framework, based on this reduction, for encoding time dependent dynamics as time independent systems. As a result, we make two advances in digital Hamiltonian simulation. First, we create a time dependent simulation algorithm based on performing qubitization on the augmented clock system, and in doing so, provide the first qubitization-based approach to time dependent Hamiltonians that goes beyond Trotterization of the ordered exponential. Second, we define a natural generalization of multiproduct formulas for time-ordered exponentials, then propose and analyze an algorithm based on these formulas. Unlike other algorithms of similar accuracy, the multiproduct approach achieves commutator scaling, meaning that this method outperforms existing methods for physically-local time dependent Hamiltonians. Our work reduces the disparity between time dependent and time independent simulation and indicates a step towards optimal quantum simulation of time dependent Hamiltonians.
What problem does this paper attempt to address?
The core problem that this paper attempts to solve is the quantum simulation problem of time - dependent Hamiltonians. Compared with time - independent Hamiltonians, the evolution operator of time - dependent Hamiltonians becomes complicated due to the existence of time ordering, which makes it difficult to directly apply time - independent simulation algorithms.
Specifically, the paper presents the following two main advances:
1. **Qubitization of time - dependent Hamiltonians based on discrete - clock construction**:
- The paper proposes a new algorithm to simulate the dynamics of time - dependent Hamiltonians by performing qubitization on an extended clock system. This is the first qubitization scheme that goes beyond the Trotterization method.
2. **Generalization of Multiproduct Formulas (MPFs)**:
- The paper defines a natural generalization of the time - ordered exponential function and proposes and analyzes a new algorithm based on these formulas. Compared with existing methods, this method shows better performance on physically local time - dependent Hamiltonians. In particular, it has the "commutator - scaling" property, that is, for time - independent and easily - simulated commutator terms, the error of this method is zero.
To achieve these goals, the paper introduces a computational framework to eliminate the influence of time ordering by encoding time - dependent dynamics into a time - independent system. The core idea of this method is to introduce a "clock degree of freedom", promote time \(t\) to a coordinate, and introduce a new evolution parameter. In this way, the dynamics of time - dependent Hamiltonians can be encoded into a time - independent Hamiltonian, thus simplifying the simulation process.
In addition, the paper also provides strict error and query complexity analysis and demonstrates the effectiveness of this method in numerical experiments. Overall, this work narrows the gap between time - dependent and time - independent Hamiltonian simulation techniques and points the way for optimal quantum simulation of time - dependent Hamiltonians.
### Key Formulas
- Evolution operator of time - dependent Hamiltonians:
\[
U(T,0) = \mathcal{T} \exp\left( -i \int_0^T H(\tau) d\tau \right)
\]
where \(\mathcal{T}\) represents the time - ordering operator.
- Floquet Hamiltonian:
\[
H_F(q, p; t, E) := H(q, p; t) - E
\]
where \(E\) is the conjugate momentum associated with time \(t\).
- Qubitized Floquet Hamiltonian:
\[
H_F = H - i \partial_t
\]
- Definition of Multiproduct Formula (MPF):
\[
U_{2,m}(t) := \sum_{j = 1}^m a_j U_2^{k_j}\left(\frac{t}{k_j}\right)
\]
where \(U_2(t)\) is the second - order Suzuki - Trotter formula, and \(a_j\) and \(k_j\) are coefficients selected through the Vandermonde linear system.
Through these methods, the paper provides a new approach for efficient quantum simulation of time - dependent Hamiltonians.