arXiv:2610.39657
Detection and Inhibition of Decoherence Through Collisions: A dynamical approach
Sagnik Chakraborty, Srijon Ghosh, Federico Amato, Gniewomir Sarbicki, Dariusz Chruściński, Rosario Lo Franco
Abstract
Quantum systems are intrinsically fragile and highly vulnerable to environmental noise, making the preservation of quantum states one of the fundamental challenges in realizing practical quantum technologies. We propose two noise-agnostic quantum state protection protocols based on a collisional model. The first protocol preserves quantum states of qubits without requiring complete prior knowledge of the underlying noise dynamics, and the second even protects a logical qubit under similar circumstances. These protocols provide robust protection for arbitrary single-qubit and entangled states against general noise processes. Furthermore, their capability to safeguard quantum states with an arbitrary number of qubits enhances the reliability, robustness, and scalability of emerging quantum architectures. These results establish the proposed framework as a promising approach for practical quantum information processing.
AI Summary
The paper studies adaptive state protection when the applied channel is unknown but belongs to a known pair $\{\Omega_1,\Omega_2\}$ with known prior probabilities. It combines noise discrimination and Petz recovery in a collisional model: a detection ancilla interacts with the system and is measured using a minimum-error discrimination measurement; a second ancilla implements the selected recovery map through a Stinespring dilation. Recovery selection follows the majority of accumulated measurement outcomes. For a known target state, the Petz map uses that state as its reference. Numerical examples cover single-qubit amplitude damping, dephasing, bit-flip and randomly generated channels, together with mixed GHZ states of two to four qubits. The reported single-qubit infidelity reaches approximately $10^{-9}$ after repeated cycles, although incorrect channel selections can cause transient increases. These results concern target-dependent recovery, rather than a common recovery operation for unknown input states. A separate protocol protects states in the repetition-code subspace $\mathcal C=\operatorname{span}\{|0\rangle^{\otimes N},|1\rangle^{\otimes N}\}$ against independent bit-flip noise. It uses a thermal reference state, weak measurements and short detection interactions. Simulations averaging over 100 Haar-random logical states report one-cycle average infidelity around $10^{-2}$ for $N=5$. With two candidate noise strengths, repeated recovery maintains fidelity above 80% initially and later approaches approximately 65%. The paper identifies measurement disturbance, longer detection interactions and greater differences between candidate noise strengths as limitations for logical-state recovery. Evidence is numerical; the construction assumes access to tailored recovery unitaries.
Central claim may not hold: The demonstrated scope is narrower than the broad protection claims: individual-state recovery requires the known target and two known candidate channels, while logical-qubit tests cover only independent bit-flip noise. The small-system examples do not establish a general guarantee of scalable protection of encoded quantum information. The stated fidelity equals Tr(rho_ini rho_reco), which gives purity rather than unity at perfect recovery of a mixed target. The specified higher-dimensional replacement-channel construction makes its Petz recovery prepare the reference state, limiting logical-state fidelity to at most one half for the stated thermal reference and contradicting the reported logical fidelities. The weak-measurement outcome probabilities and conditional state updates do not describe the same quantum measurement instrument.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 9 pages, 6 figures