| Quantum Phase Estimation | |
|---|---|
| Name | Quantum Phase Estimation |
| Synonyms | QPE |
| Inventors | Peter Shor (in part), Alexei Kitaev (Kitaev's approach) |
| Introduced | 1990s |
| Field | Quantum computing / Quantum information theory |
| Hardware | Quantum circuit model, Ion trap, Superconducting qubit, Photonic quantum computing |
Quantum Phase Estimation
Quantum Phase Estimation (QPE) is a quantum algorithm that estimates the eigenphase (phase) φ associated with an eigenstate of a unitary operator. It underpins many quantum algorithms for problems in quantum simulation, quantum chemistry, and number theory, enabling exponential speedups over classical methods for tasks such as eigenvalue estimation and order finding. QPE is central to the quantum algorithm toolkit because accurate phase information maps directly to energy spectra and periodicity properties used in algorithms like Shor's algorithm.
Quantum Phase Estimation provides a protocol to extract spectral information of a unitary evolution U by mapping eigenvalues e^{2πiφ} to measurable bit strings. In quantum mechanics, phases encode observable quantities such as energy via time evolution under a Hamiltonian H: U = e^{-iHt/ħ}. Thus QPE connects the abstract quantum circuit model to physical quantities, enabling tasks in quantum simulation and precision quantum metrology. The algorithm is considered a cornerstone for algorithms that exploit quantum parallelism and interference, and it is an enabling primitive for quantum algorithms developed by researchers at institutions such as IBM, Google, and academic groups at MIT and Caltech.
Given a unitary operator U acting on a Hilbert space and an eigenstate |ψ⟩ with U|ψ⟩ = e^{2πiφ}|ψ⟩, the problem is to estimate the phase φ ∈ [0,1) to m bits of precision. QPE constructs a composite system with an ancilla register of t qubits prepared in superposition via Hadamard gates and a register holding |ψ⟩. Controlled-U^{2^j} operations imprint powers of the phase onto the ancilla. A quantum Fourier transform (QFT) on the ancilla followed by projective measurement yields an integer estimate k such that k/2^t ≈ φ. The algorithm's error can be bounded in terms of t, the number of ancilla qubits, and the fidelity of state preparation and controlled evolutions.
The canonical circuit uses three stages: state preparation of |ψ⟩, phase kickback via controlled-U^{2^j} gates, and an inverse inverse QFT on the ancilla register. The circuit leverages the phase kickback phenomenon whereby eigenphases translate into relative phases on control qubits. Implementation requires modular exponentiation techniques in contexts like Shor's algorithm or Hamiltonian simulation methods (e.g., Trotter–Suzuki decomposition, qubitization, Linear Combination of Unitaries). The circuit depth, gate count, and required ancilla are primary resources influencing practical feasibility on platforms like superconducting qubit processors or ion trap systems.
Several variants trade qubit count against circuit depth and classical postprocessing: - Kitaev's algorithm uses single control qubits and adaptive measurements to estimate phase bits sequentially, reducing ancilla requirements at the cost of more rounds and classical control. - Iterative QPE (IQPE) implements bit-by-bit estimation using feedback and phase shifts, enabling estimation with one ancilla qubit. - Bayesian phase estimation applies statistical inference and adaptive measurement scheduling to improve precision with fewer repetitions; it is closely related to techniques in quantum metrology and adaptive measurement protocols. Resource analysis considers queries to controlled-U, coherence time, and gate fidelity. Complexity statements often express precision ε dependence: O(1/ε) controlled-U queries for standard formulations, with variants achieving near-optimal scaling in specific regimes.
QPE is applied to: - Estimating molecular energies in quantum chemistry via simulation of electronic structure Hamiltonians (used in algorithms proposed by Aspuru-Guzik et al.). - Determining eigenvalues of large sparse matrices arising in condensed-matter physics and lattice models using simulation primitives developed at labs like Los Alamos National Laboratory and in projects such as Quantum Simulation initiatives. - Enabling Shor's algorithm by providing period finding as phase estimation of modular multiplication. - Improving parameter estimation and frequency measurements in quantum metrology, where phase sensitivity is paramount; connections exist to Heisenberg limit scaling and entangled probe states such as NOON states.
Practical implementations must manage errors from Trotterization, approximation of time evolution, gate infidelity, and decoherence. Error models include stochastic noise, coherent control errors, and readout errors; error mitigation techniques such as zero-noise extrapolation, digital error correction using quantum error correction codes, and randomized compiling can improve accuracy. Finite-sampling statistics impose shot-noise limits; combining QPE with amplitude amplification or improved state preparation affects success probability. Resource estimates typically balance ancilla count, total circuit depth, and controlled-U repetition to achieve target precision under hardware constraints.
QPE and its variants have been demonstrated on multiple platforms: small-scale implementations on trapped ion systems (e.g., experiments at National Institute of Standards and Technology groups), superconducting qubit processors (notably by teams at IBM and Google), and photonic setups using linear optics. Proof-of-principle demonstrations often estimate low-bit phases or molecular eigenvalues for minimal basis models. Ongoing experimental efforts focus on integrating improved Hamiltonian simulation techniques, scalable control for controlled-U^{2^j} operations, and incorporation of quantum error correction to extend coherence for deeper phase-estimation circuits.
Category:Quantum algorithms Category:Quantum information theory