Petz recovery maps of single-qubit decoherence channels in an ion trap quantum processor
Wen-Han Png 1, Valerio Scarani 1,2 · 2025
Read Petz recovery maps of single-qubit decoherence channels in an ion trap quantum processor by Wen-Han Png 1, Valerio Scarani 1,2 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-20399.txt.
Introduction
Wen-Han Png and Valerio Scarani present a 2025 paper on a near best method for undoing noise in a single quantum bit, built for an ion trap machine. Quantum noise scrambles stored states, and full recovery is impossible. The Petz map is a known recipe that comes close to the best reversal for a given noise type.
The authors design circuits that run the exact map for a chosen starting state. The circuits need at most one or two helper bits and a modest gate count. They cover three standard noise types and model real trap flaws in simulation. They also state how exact the prior knowledge of the state must be.
Catalog record
| CREATOR | Wen-Han Png 1, Valerio Scarani 1,2 |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 46 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-20399
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Petz recovery maps of single-qubit decoherence channels in an ion trap quantum processor Wen-Han Png 1, Valerio Scarani 1,2 Abstract The Petz recovery map provides a near-optimal reversal of quantum noise, yet proposals for its implementation are only recent. We propose a physical realization of the exact state-specific Petz map in an ion trap for qubit decoherence channels. Our circuit constructions require at most 1 ( 2 ) 1(2) ancilla qubits and 3 ( 20 ) 3(20) CNOT gates for channels with Kraus rank 2 ( > 2 ) 2(>2) . We analyze typical ion trap errors and construct corresponding Petz maps, simulating their performance under realistic noise modeled by residual spin-motion coupling. Quantum circuits are provided for depolarizing, dephasing, and amplitude damping channels. Focusing on single-shot recovery, suited for present-day devices, we also quantify the precision of prior knowledge required to achieve a recovery error below 0.01 across varying decoherence levels and state purities. Preprint: APS/123-QED I Introduction Quantum systems are never truly isolated. Processes and protocols are done in the presence of environments that introduce noise. As such, recovery protocols that can undo the unintended dynamics are of great utility. Petz’s theory of statistical sufficiency [ 1 ] defines a transpose channel \mathcal{R}_{(\mathcal{E},\gamma)} (also known as Petz recovery map, defined for a quantum channel \mathcal{E} and reference state \gamma ) that recover a quantum state affected by noise [ 2 , 3 ] . The Petz map can be understood as a quantum generalization of Bayes’ theorem [ 4 , 5 , 6 , 7 ] . As a brief intorduction, Bayes’ theorem updates a prior distribution \gamma(\theta) over parameters \theta based on observed data x , with the posterior distribution given by: \hat{\varphi}_{\gamma}(\theta|x)=\frac{\varphi(x|\theta)\gamma(\theta)}{\varphi[\gamma](x)} (1) Here, \varphi(x|\theta) is the likelihood, and \varphi[\gamma](x)=\sum_{\theta}\varphi(x|\theta)\gamma(\theta) is the normalization factor. From an information-theoretic perspective, Bayes’ theorem acts as a reversal of the likelihood function \varphi(x|\theta) , reconstructing the distribution over \theta based on given data. As we shall describe in detail later, the Petz recovery map \mathcal{R}_{\mathcal{E},\gamma} serves as the quantum analog, where the likelihood corresponds to a quantum channel \mathcal{E} and the prior is represented by a density matrix \gamma . Here, \gamma is also often referred to as a reference state. It is not a physical quantum state, rather it represents the prior knowledge we have for the input state before applied with \mathcal{E} . It is also the state that we choose to recover perfectly by applying the Petz map on the output of the channel: indeed, \mathcal{R}_{\mathcal{E},\gamma}[\mathcal{E}[\gamma]]=\gamma . The Petz recovery map is often used in quantum information as a theoretical tool, primarily for checking to what extent recovery from noise is possible. The performance, dependence on reference state, and dimensionality of the Petz recovery map has been investigated for typical quantum channels such as dephasing and amplitude damping [ 8 ] . Notably, it is known that the map is near-optimal in reversing the effect of noise [ 9 ] . It has also been shown as a universal recovery operation for approximate quantum error correction, which is once again near-optimal in terms of worst-case fidelity [ 10 , 11 , 12 ] . It has also been applied in the decoding process to obtain a coherent information rate in quantum communication channels [ 13 ] , with further applications in quantum channel capacities [ 14 , 15 , 16 , 17 ] , state discrimination [ 9 , 18 , 19 , 20 ] , and entanglement wedge reconstruction [ 21 ] . Recent studies have proposed circuit implementations of the Petz map. In the first [ 22 ] , the state-specific Petz map [ 9 ] was described using block encoding and quantum singular value transformation. An exact implementation of the code-specific Petz map [ 10 ] was then proposed in [ 23 ] . There, the authors exploited the fact that the Petz map is a CPTP map, and as such can be isometrically extended to a unitary circuit by Stinespring dilation. Besides being exact, the implementation of the isometric extension is also better in its scaling with the number of qubits to be corrected. In this work, we take a further step towards practicality and describe the implementation of Petz recovery maps for a trapped-ion qubit and its most frequent decoherence channels. The map will be implemented using geometric phase gates as the interaction with the ancilla. We analyze the error thresholds required for the geometric phase gate to ensure reliable operation. This paper is structured as follows, In Sec. II we introduce the Petz recovery map and its circuit implementation for arbitrary-rank decoherence channel affecting a single qubit; then we examine the requirements for the prior distribution of the reference state to ensure that the recovery error is bounded below 0.01. In Sec. III , we describe the trapped ion system and common error sources that give rise to dephasing, amplitude damping, and depolarization. For each of these errors, we determine the error thresholds of the geometric phase gate needed to realize the corresponding Petz recovery maps. II Petz recovery map In this section, we briefly introduce the Petz recovery map. To begin with, we consider a forward channel \mathcal{E} acting on a qubit input, where \mathcal{E} is a CPTP map of rank M given by \mathcal{E}(\bullet)=\sum_{m=1}^{M}E_{m}\bullet E_{m}^{{\dagger}} . For a qubit channel, \bullet\in\mathbb{C}^{2} and M={1,...,4} . One can regard \mathcal{E} as a channel that introduces undesired dynamics, commonly regarded as noise or decoherence. The question of whether the noise can be reversed was first addressed in [ 1 ] . As a consequence of statistical sufficiency in Petz’s theory, a decohered state \mathcal{E}(\rho) can be fully recovered, e.g. \hat{\mathcal{R}}_{(\mathcal{E},\gamma)}\circ\mathcal{E}(\rho)=\rho if and only if D(\mathcal{E}(\rho)|\mathcal{E}(\gamma))=D(\rho|\gamma) (2) where D(\rho|\gamma)=\text{Tr}[\rho(\log\rho-\log\gamma)] is the Umegaki quantum relative entropy of the two states. Given \mathcal{E} and a reference state \gamma , the corresponding transpose channel of \mathcal{E} can be constructed as follows \hat{\mathcal{R}}_{(\mathcal{E},\gamma)}(\bullet)=\sqrt{\gamma}\mathcal{E}^{\dagger}\left(\frac{1}{\sqrt{\mathcal{E}(\gamma)}}\bullet\frac{1}{\sqrt{\mathcal{E}(\gamma)}}\right)\sqrt{\gamma} (3) In Kraus representation, the Petz map read \hat{\mathcal{R}}_{(\mathcal{E},\gamma)}(\bullet)=\sum_{m=1}^{M}K_{m}^{(\mathcal{E},\gamma)}\bullet K_{m}^{{\dagger}(\mathcal{E},\gamma)} (4) where K_{m}^{(\mathcal{E},\gamma)}=\sqrt{\gamma}E_{m}^{\dagger}\frac{1}{\sqrt{\mathcal{E}(\gamma)}} . The recovery fidelity of Petz map is given by \mathcal{F}(\rho,\hat{\mathcal{R}}_{(\mathcal{E},\gamma)}\circ\mathcal{E}(\rho)) , where the fidelity is defined as \mathcal{F}(\rho_{0},\rho_{1})=\text{Tr}\Big[\sqrt{\sqrt{\rho_{0}}\rho_{1}\sqrt{\rho_{0}}}\Big]^{2} . II.1 Isometric extension for the Petz Map of a single-qubit decoherence channel A Petz recovery map is a CPTP map in general \hat{\mathcal{R}}_{(\mathcal{E},\gamma)}:\mathcal{S}(\mathcal{H}_{A})\rightarrow\mathcal{S}(\mathcal{H}_{C}) . We define the ancillary system as \rho_{B}\in\mathcal{S}(\mathcal{H}_{B}) where \mathcal{H}_{B} is the Hilbert space of the ancillary system. The isometric extension of the \hat{\mathcal{R}}_{(\mathcal{E},\gamma)} can be constructed by defining a unitary channel U_{(\mathcal{E},\gamma)}\bullet U_{(\mathcal{E},\gamma)}^{\dagger}:\mathcal{S}(\mathcal{H}_{B}\otimes\mathcal{H}_{A})\rightarrow\mathcal{S}(\mathcal{K}),\mathcal{K}=\mathbb{C}^{2M} , followed by tracing the ancillary system \hat{\mathcal{R}}_{(\mathcal{E},\gamma)}(\bullet)=\text{Tr}_{B}\left[U_{(\mathcal{E},\gamma)}(\rho_{B}\otimes\bullet)U_{(\mathcal{E},\gamma)}^{{\dagger}}\right]. (5) Using the trace-preserving (TP) condition \sum_{m=1}^{M}K_{m}^{{\dagger}(\mathcal{E},\gamma)}K_{m}^{(\mathcal{E},\gamma)}=\mathcal{I}_{A} , the sum of each diagonal entry gives 1 each, while the sum of each off-diagonal term gives zero. Let \rho_{B}=|0\rangle\langle 0| , the TP condition allows us to construct U_{(\mathcal{E},\gamma)} using Stinespring’s dilation U_{(\mathcal{E},\gamma)}=\left[\begin{array}[]{cccc}K_{1}^{(\mathcal{E},\gamma)}&.&.&.\\ K_{2}^{(\mathcal{E},\gamma)}&.&.&.\\ \vdots&.&.&.\\ K_{M}^{(\mathcal{E},\gamma)}&.&.&.\end{array}\right] (6) where the first two columns are mutually orthogonal. Here, U_{(\mathcal{E},\gamma)}\in U(2M)/U(1) up to a global phase. The rest of the column of U_{(\mathcal{E},\gamma)} can be constructed using Gram-Schmidt orthogonalisation. Next, we discuss the resources required for circuit implementation of U_{(\mathcal{E},\gamma)} in terms of number of ancillary qubit and number of CNOT gates. For the rest of the paper, we only consider the Petz recovery map for single-qubit decoherence.