[Paper Review] Quantum Self-Correcting Stabilizer Codes

Alastair Kay, Roger Colbeck|ArXiv.org|Oct 20, 2008

Quantum Computing Algorithms and Architecture19 citations

TL;DR

This paper proves that 2D quantum stabilizer Hamiltonians on qubit lattices cannot achieve self-correction for quantum information storage, as local thermal noise can induce logical errors via explicit string-like operators that loop around the lattice. The authors construct such operators for any stabilizer code, showing their survival time against thermal noise is not exponentially protected, thus ruling out these systems as candidates for a quantum hard drive.

ABSTRACT

In this paper, we explicitly construct (Abelian) anyonic excitations of arbitrary stabilizer Hamiltonians which are local on a 2D lattice of qubits. This leads directly to the conclusion that, in the presence of local thermal noise, such systems cannot be used for the fault-tolerant storage of quantum information by self-correction i.e. they are ruled out as candidates for a `quantum hard drive'. We suggest that in 3D, the same construction leads to an argument that self-correction is impossible.

Motivation & Objective

• To determine whether topological quantum stabilizer codes in 2D can achieve self-correction against local thermal noise.
• To investigate whether the absence of self-correction in the toric code extends to all stabilizer Hamiltonians on 2D lattices.
• To explore whether similar no-go arguments apply in 3D and higher spatial dimensions.
• To provide a constructive method for identifying error paths that destabilize quantum information in such systems.

Proposed method

• Constructs explicit Abelian anyonic excitations (string operators) for arbitrary stabilizer Hamiltonians on 2D qubit lattices.
• Demonstrates that these string operators act non-trivially on the ground state space and commute with the Hamiltonian terms.
• Uses truncation of string operators to generate excited states with finite energy cost, modeling thermal noise effects.
• Applies Corollary 2 and Lemma 9 to show independence of degeneracy-breaking operators across elementary sets.
• Extends the 2D construction to d-dimensional lattices by replacing loops with (d−1)-dimensional objects, such as area operators in 3D.
• Analyzes logical gate structures and anti-commuting terms to assess protection mechanisms in higher dimensions.

Experimental results

Research questions

• RQ1 Can 2D stabilizer Hamiltonians on qubit lattices be self-correcting against local thermal noise?
• RQ2 Do all stabilizer codes in 2D admit string-like operators that provide paths for logical errors under thermal noise?
• RQ3 Is the absence of self-correction in the toric code a generic feature of all 2D stabilizer codes?
• RQ4 Can the same no-go argument be extended to 3D stabilizer Hamiltonians?
• RQ5 Under what conditions in 4D might self-correction be possible, given the failure in 2D and 3D?

Key findings

• All 2D stabilizer Hamiltonians with periodic boundary conditions admit string operators that loop around the torus and act non-trivially on the ground state space.
• These string operators provide explicit paths for local thermal noise to induce logical errors, preventing exponential suppression of error rates.
• The survival time of quantum information against thermal noise is not exponentially increased with system size, ruling out self-correction in 2D.
• In 3D, the construction suggests the existence of similar error paths, implying self-correction is also impossible for stabilizer Hamiltonians in 3D.
• In 4D, the structure of stabilizer terms allows for 2D logical operations that could provide protection analogous to the 2D Ising model, suggesting a potential path to self-correction.
• The anyonic excitations formed by truncating string operators behave like anyons with Abelian braiding statistics, identical to those in the toric code but with up to |G̃| distinct particle types.