Surface Code Error Correction with Crosstalk Noise

Zeyuan Zhou, Andrew Ji, Yongshan Ding · 2025

Read Surface Code Error Correction with Crosstalk Noise by Zeyuan Zhou, Andrew Ji, Yongshan Ding (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-2503-04642.txt.

Introduction

Zeyuan Zhou, Andrew Ji, and Yongshan Ding study, in a 2025 paper, how crosstalk noise strains the surface code, a leading design for quantum error correction. Error correction theory usually assumes noise that is independent and forgetful, because that is easy to simulate. Real hardware misbehaves in a correlated way: operations leak into neighbors, including the helper qubits meant to catch errors.

The authors systematically vary crosstalk types and measure the damage in memory and stability experiments. Two defenses follow from the findings. Flag qubits can catch the dangerous correlated faults, and redundant stabilizer checks add extra witnesses. Numerical and analytical work together test both defenses. The team also studies logical crosstalk inside a code block storing several qubits, and derives conditions under which physical crosstalk stays harmless. The results guide building codes that survive the noise chips actually make.

Catalog record

CREATORZeyuan Zhou, Andrew Ji, Yongshan Ding
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE38 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-04642

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Surface Code Error Correction with Crosstalk Noise Zeyuan Zhou, Andrew Ji, Yongshan Ding Abstract The design and performance analysis of quantum error correction (QEC) codes are often based on incoherent and independent noise models since it is easy to simulate. However, these models fail to capture realistic hardware noise sources, such as correlated errors (crosstalk), which can significantly impact QEC code performance, especially when they occur between data and ancillary qubits. In this paper, we systematically study various types of crosstalk noise and quantify their effects on surface codes through memory and stability experiments. Based on our findings, we introduce crosstalk-robust implementations of QEC via flag qubit designs and redundant stabilizer checks. We perform both numerical and analytical studies to demonstrate the efficacy of these strategies. In addition, we analyze logical crosstalk in an [ [ n , k > 1 , d ] ] [[n,k>1,d]] code block and establish analytical conditions under which physical crosstalk does not lead to logical crosstalk. Together, our analytical and numerical results shed light on designing QEC codes that are robust against hardware realistic crosstalk noise, paving the way for reliable experimental realization of fault-tolerant quantum computing. I Introduction Quantum error correction (QEC) emerges as a crucial component to realize large-scale fault-tolerant quantum computing (FTQC) [ 1 , 2 ] . QEC encodes quantum information into a larger Hilbert space, which allows for detection and correction via repetitive measurements of its stabilizers. Traditionally, QEC studies assume an independent and incoherent noise model for stabilizer design and threshold simulation [ 1 ] . This model, however, falls short in capturing all error properties of a physical system. For example, correlated error, also known as crosstalk, is a detrimental realistic noise on hardware. Generally characterized as an unwanted interaction, crosstalk takes various forms in hardware architectures. For example, in superconducting and semiconductor platforms, qubits are susceptible to parasitic ZZ interaction from always-on coupling between neighboring qubits [ 3 , 4 , 5 , 6 ] . In trapped-ion systems, unwanted XX -type interactions are commonly seen in the implementation of multi-qubit gates, such as the Mølmer–Sørensen gate [ 7 , 8 , 9 ] . Neutral atom systems also experience crosstalk when atom blockades are moved together to implement entangling gates [ 10 , 11 ] . While crosstalk suppression has been thoroughly studied from both the hardware-level (e.g. coupler and gate design [ 12 , 13 , 14 ] ) and the software-level (e.g. compiling/mapping [ 15 , 16 , 17 , 18 , 19 , 20 , 21 ] , pulse design [ 22 , 23 , 24 , 25 , 26 , 27 ] ), the investigation of its impact on QEC [ 8 , 28 , 29 , 30 , 31 , 32 , 33 , 34 ] is still limited to certain platform-specific crosstalk type and lacks general measures to address crosstalk when instantiating a QEC code. Here, we provide a comprehensive analysis of the robustness of surface code, one of the most practical and scalable QEC codes, towards various types of crosstalk. Specifically, we perform memory and stability experiments for each code under a “depolarizing + crosstalk” noise model, from which we find the effect of crosstalk on both state preservation and fault-tolerant logical operation. Both crosstalk and depolarizing noise models are circuit-level and are inspired by state-of-the-art qubits calibration and device connectivity data. Specifically, we survey through four types of crosstalk on two dimensions: gate-based or always-on, and between data-data qubits or between data-ancilla qubits. Numerical results show that among all four types, gate-based crosstalk between data and ancilla qubits poses the most detrimental effect on both memory and FTQC. To address this issue, we propose two schemes: a flagged syndrome extraction design for surface code and a redundant stabilizer check method suitable for repetition code. Both measures enable effective detection of crosstalk and are hardware agnostic, which means gate optimization or engineering is not required. The schemes are also generalizable to more codes. We perform both numerical and analytical studies to quantify the efficacy of these strategies and identify the noise regime where those techniques work the best. Apart from physical crosstalk on [[n,1,d]] codes, we also introduce for the first time “logical crosstalk” in an [[n,k>1,d]] code block. We show that physical crosstalk can be converted to logical correlated errors contingent on the definition of logical Pauli operators. We show how logical crosstalk could impact logical gates using [[4,2,2]] and as an example [ 35 , 36 ] and calculated the effective logical crosstalk error rate under different physical noise model. Besides, we develop a crosstalk conversion condition and show analytically that on a 2 D sparse connectivity, low-weight logical crosstalk will not happen with large d . Together, these results fill the gap in the studies of how crosstalk impacts QEC and provide novel approaches to enhance crosstalk robustness. II Noise Model We simulate a circuit-level noise model that includes both depolarizing and crosstalk noise. In this model, depolarizing noise serves as the baseline, meaning that after each logical gate or reset operation, an incoherent Pauli error is inserted with a probability controlled by a single parameter p . Formally, the depolarizing noise channel is expressed as: \displaystyle\mathcal{E}_{1}(\rho) \displaystyle= \displaystyle(1-p_{1})\rho+\frac{p_{1}}{3}\sum_{i}\sigma^{i}\rho\,\sigma^{i}, (1) \displaystyle\mathcal{E}_{2}(\rho) \displaystyle= \displaystyle(1-p_{2})\rho+\frac{p_{2}}{15}\sum_{i,j}\bigl[\sigma^{i}\otimes\sigma^{j}\bigr]\rho\bigl[\sigma^{i}\otimes\sigma^{j}\bigr], (2) where i,j\in\{I,X,Y,Z\} and (i,j)\neq(I,I) . We parametrize all error probabilities with the single parameter p and scale it according to different gate scenarios. When conducting experiments, we vary p to observe how the logical error rates respond. The specific scaling factors for p in different gates are inspired by hardware characterization data across multiple quantum computing platforms. Beyond depolarizing noise, we also define crosstalk noise as a correlated N -body Pauli error channel: \mathcal{E}_{N}(\rho)=(1-p_{X})\,\rho+p_{X}\,\hat{P}\,\rho\,\hat{P},\quad\hat{P}=\hat{\sigma}_{i}^{\otimes N}, (3) where i\in\{X,Y,Z\} . In particular, we consider four classes of crosstalk (see Table II ) that are highly relevant to existing quantum hardware platforms. For example, in superconducting tunable qubit architectures, ZZ -type crosstalk typically dominates during two-qubit gates. Meanwhile, in fixed-frequency transmon qubits, crosstalk often appears as always-on ZZ coupling between nearest neighbors (N.N.) [ 4 , 37 , 38 ] . Similarly, in trapped-ion systems, residual XX -type coupling exists when implementing two-qubit gates [ 7 , 8 , 9 ] . We categorize these crosstalk interactions as either gate-based or always-on , and they can occur between data and ancilla qubits or among data qubits themselves. In a 2D grid layout for syndrome extraction, data and ancilla qubits are typically placed as nearest neighbors to reduce routing overhead, hence enabling such crosstalk pathways. Additionally, data qubits can experience further crosstalk from next-nearest-neighbor (N.N.N.) interactions or multi-qubit gates. Notably, N.N.N. ZZ -type coupling is common in superconducting systems [ 39 , 40 ] , though its strength can be orders of magnitude lower than N.N. coupling. Multi-qubit interactions frequently occur in neutral atom platforms where atom blockades are brought together for multi-qubit gates [ 10 , 11 ] . For always-on crosstalk , we parametrize the crosstalk error through a Pauli-twirling approximation (PTA) as follows: \displaystyle p_{ZZ}(t)\;=\;\sin^{2}\bigl(J_{ZZ}\,t\bigr), (4) where the coefficient J_{ZZ} represents the coupling strength. Here, we use publicly available IBM Quantum device data to set J_{ZZ}=10 kHz [ 24 , 27

Read the complete text