Quantum dual extended Hamming code immune to collective coherent errors
En-Jui Chang · 2025
Read Quantum dual extended Hamming code immune to collective coherent errors by En-Jui Chang (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-2503-05249.txt.
Introduction
En-Jui Chang constructs a family of quantum error codes in a 2025 paper that resist a shared drift error. All qubits in a device drift together under their own physics, and that joint drift can damage stored data more than random single errors. Long storage makes the joint drift the main threat, as in trapped ion systems.
Chang builds codes that hold a fixed total excitation, which blocks the joint drift by design. The codes keep a high data rate while still fixing small errors. The smallest code stores one logical qubit in eight physical ones. It roughly doubles the data rate of the best past code of this type and lifts its error threshold about four fold.
Catalog record
| CREATOR | En-Jui Chang |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 20 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-05249
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Quantum dual extended Hamming code immune to collective coherent errors En-Jui Chang Abstract Collective coherent (CC) errors are inevitable, as every physical qubit undergoes free evolution under its kinetic Hamiltonian. These errors can be more damaging than stochastic Pauli errors because they affect all qubits coherently, resulting in high-weight errors that standard quantum error-correcting (QEC) codes struggle to correct. In quantum memories and communication systems, especially when storage durations are long, CC errors often dominate over stochastic noise. Trapped-ion platforms, for example, exhibit strong CC errors with minimal stochastic Pauli components. In this work, we address the regime where immunity to CC errors, high code rate (due to limited qubit availability), and moderate distance (sufficient for correcting low-weight errors) are all essential. We construct a family of constant-excitation (CE) stabilizer codes with parameters [ [ 2 r + 1 , 2 r − ( r + 1 ) , 3 ] ] [[2^{r+1},2^{r}-(r+1),3]] . The smallest instance, the [ [ 8 , 1 , 3 ] ] [[8,1,3]] code, improves the code rate and error threshold of the best previously known CE code by factors of approximately two and four, respectively. I Introduction The intrinsic Hamiltonian governing each qubit’s free evolution drives coherent time dynamics, even when the qubits are idle or operating within well-controlled quantum platforms. As a result, collective coherent (CC) errors are unavoidable. Unlike stochastic Pauli errors, CC errors apply coherent, high-weight operations simultaneously across many qubits, potentially causing more severe logical errors. Although these errors do not reduce the state’s purity, their structured nature makes them particularly damaging to quantum information. To solve this issue, significant progress has been made in correcting these CC errors [ 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ] . Although quantum memory, communication, and computation differ fundamentally from their classical counterparts, they are likely to share some common practical considerations. First, multiple sources of qubits will exist, offering varying levels of quality and cost. Second, quantum data will need to be stored for future use, often for uncertain durations. Third, quantum information may be transmitted to mobile or spatially distributed receivers whose precise locations are not always known. Fourth, transferring quantum data from memory to processing units incurs time overhead. We will address these considerations in detail later in this work. Trapped-ion qubits [ 10 , 11 , 12 , 13 , 14 , 15 ] share architectural similarities with classical Von Neumann systems, where memory and computation are physically separated. In such platforms, qubit transport, whether through optical tweezers or ion shuttling, naturally introduces CC errors. While trapped-ion systems are known for their high-fidelity operations, they typically suffer from slower gate speeds and a limited maximum qubit count due to their inherently one-dimensional layout. As a result, the primary bottleneck lies in the number of high-quality qubits that can be reliably operated within a single device. In addition to CC errors, quantum systems are also subject to depolarizing noise, typically modeled as stochastic Pauli errors. Two well-known quantum error-correcting (QEC) codes that address such errors are the Shor code [ 16 ] and the Steane code [ 17 , 18 ] . Together, they form the class of Calderbank-Shor-Steane (CSS) stabilizer codes. The Shor code offers a systematic way to increase the code distance, a measure of error-correction strength, at the expense of code rate, defined as the ratio of logical qubits to physical qubits. In contrast, the Steane code demonstrates how to improve the code rate while fixing the code distance. These two design philosophies highlight a broader principle: effective QEC must adapt to varying hardware constraints, including the availability and quality of physical qubits. To address both CC and stochastic Pauli errors, QEC codes with constant excitation (CE) codewords are essential for reliable quantum memory and computation. In our previous work [ 8 ] , we made significant progress in increasing the code distance without sacrificing too much code rate. The family of [[2(w+1)(w+K),K]] CE codes, referred to as w -codes, can be interpreted as [[2(w+1)(w+K),K,d^{\prime}]] Shor-like codes with distance d^{\prime}=w+1 for K=1 , d^{\prime}=2 for K\geq 2 . Here, the parameter w indicates the ability to correct weight- w amplitude-damping (AD) errors, though not weight- w Pauli errors. For conventional [[n,k,d]] stabilizer codes designed to correct Pauli errors, up to \lfloor\frac{d}{2}\rfloor errors can be detected and \lfloor\frac{d-1}{2}\rfloor errors can be corrected. In contrast, the present work prioritizes code rate over distance, targeting scenarios where only a limited number of high-quality qubits are available and where correcting weight-1 Pauli errors is sufficient. These qubits may serve as quantum memory or as carriers in quantum communication, where timing information is often inaccessible, making it impractical to correct CC errors simply by reversing coherent evolution. As a result, we aim to design Steane-like CE codes optimized for memory and communication applications, accepting reduced flexibility in code distance in exchange for higher encoding efficiency. Before constructing our CE codes, we first examine how existing CE codes are designed. The simplest example is the [[2,1,0]] complemented dual-rail (CDR) code [ 19 ] , which provides no error protection. A major class of CE codes, including our previous work [ 8 ] , is based on concatenating a standard outer stabilizer code with an inner CDR code. This approach underlies several known constructions, such as the [[8,1,2]] [ 20 ] , [[18,1,3]] , and [[32,1,4]] CE codes, each requiring twice the number of qubits compared to their non-CE stabilizer counterparts. Movassagh and Ouyang [ 21 ] proposed the [[15,1,4]] CE code by extending beyond the stabilizer framework. While it satisfies the QEC conditions, it lacks a clear method for syndrome extraction and conditional recovery operations. Another example, the [[5,1,2]] CE code described as Example 6 in [ 4 ] , appears distinct but is effectively a [[4,1,2]] CE code that still relies on concatenation with the inner CDR code. From these observations, especially from [ 4 ] , we recognize that the inner CDR code effectively acts as a two-qubit repetition code, which increases the overall code distance relative to the outer code. This suggests that many prior constructions rely on overly strong outer codes, potentially limiting code rate unnecessarily. As a result, we propose designing an outer code that does not differentiate between X and Z errors, relying instead on the inner CDR code to enable effective discrimination between these error types. In this work, we introduce a new family of CE stabilizer codes with parameters [[2^{r+1},2^{r}-(r+1),3]] . These codes are constructed by encoding classical extended Hamming codes [2^{r},2^{r}-(r+1),4] [ 22 ] into CE subspaces using an inner CDR code. The smallest instance in this family is the [[8,1,3]] code, which uses fewer qubits than previously known CE codes, whether stabilizer or non-stabilizer. Moreover, this new construction improves the asymptotic code rate to \frac{1}{2} . A summary comparison is provided in Table 1 . Table 1: Comparison of representative CE QEC codes. Each [[n,k,d]] code encodes k logical qubits into n physical qubits with distance d , allowing correction of up to \lfloor\frac{d-1}{2}\rfloor and detection of up to \lfloor\frac{d}{2}\rfloor Pauli errors. The rate is defined as the ratio \frac{k}{n} , indicating