Re-locative guided search optimized self-sparse attention enabled deep learning decoder for quantum error correction
Abstract Heavy hexagonal coding is a type of quantum error-correcting coding in which the edges and vertices of a low-degree graph are assigned auxiliary and physical qubits. While many topological code decoders have been presented, it is still difficult to construct the optimal decoder due to leaka...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-01-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-87782-2 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832571723634966528 |
|---|---|
| author | Umesh Uttamrao Shinde Ravikumar Bandaru |
| author_facet | Umesh Uttamrao Shinde Ravikumar Bandaru |
| author_sort | Umesh Uttamrao Shinde |
| collection | DOAJ |
| description | Abstract Heavy hexagonal coding is a type of quantum error-correcting coding in which the edges and vertices of a low-degree graph are assigned auxiliary and physical qubits. While many topological code decoders have been presented, it is still difficult to construct the optimal decoder due to leakage errors and qubit collision. Therefore, this research proposes a Re-locative Guided Search optimized self-sparse attention-enabled convolutional Neural Network with Long Short-Term Memory (RlGS2-DCNTM) for performing effective error correction in quantum codes. The integration of the self-sparse attention mechanism in the proposed model increases the feature learning ability of the model to selectively focus on informative regions of the input codes. In addition, the use of statistical features computes the statistical properties of the input, thus aiding the model to perform complex tasks effectively. For model tuning, this research utilizes the RIGS nature-inspired algorithm that mimics the re-locative, foraging, and hunting strategies, which avoids local optima problems and improves the convergence speed of the RlGS2-DCNTM for Quantum error correction. When compared with other methods, the proposed RlGS2-DCNTM algorithm offers superior efficacy with a Minimum Mean Squared Error (MSE) of 4.26, Root Mean Squared Error of 2.06, Mean Absolute Error of 1.14 and a maximum correlation and $$R^2$$ of 0.96 and 0.92 respectively, which shows that the proposed model is highly suitable for real-time error decoding tasks. |
| format | Article |
| id | doaj-art-58573fd2bacc481db21ad3ee8f64114e |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-58573fd2bacc481db21ad3ee8f64114e2025-02-02T12:22:10ZengNature PortfolioScientific Reports2045-23222025-01-0115111910.1038/s41598-025-87782-2Re-locative guided search optimized self-sparse attention enabled deep learning decoder for quantum error correctionUmesh Uttamrao Shinde0Ravikumar Bandaru1Department of Mathematics, School of Advanced Sciences, VIT-AP UniversityDepartment of Mathematics, School of Advanced Sciences, VIT-AP UniversityAbstract Heavy hexagonal coding is a type of quantum error-correcting coding in which the edges and vertices of a low-degree graph are assigned auxiliary and physical qubits. While many topological code decoders have been presented, it is still difficult to construct the optimal decoder due to leakage errors and qubit collision. Therefore, this research proposes a Re-locative Guided Search optimized self-sparse attention-enabled convolutional Neural Network with Long Short-Term Memory (RlGS2-DCNTM) for performing effective error correction in quantum codes. The integration of the self-sparse attention mechanism in the proposed model increases the feature learning ability of the model to selectively focus on informative regions of the input codes. In addition, the use of statistical features computes the statistical properties of the input, thus aiding the model to perform complex tasks effectively. For model tuning, this research utilizes the RIGS nature-inspired algorithm that mimics the re-locative, foraging, and hunting strategies, which avoids local optima problems and improves the convergence speed of the RlGS2-DCNTM for Quantum error correction. When compared with other methods, the proposed RlGS2-DCNTM algorithm offers superior efficacy with a Minimum Mean Squared Error (MSE) of 4.26, Root Mean Squared Error of 2.06, Mean Absolute Error of 1.14 and a maximum correlation and $$R^2$$ of 0.96 and 0.92 respectively, which shows that the proposed model is highly suitable for real-time error decoding tasks.https://doi.org/10.1038/s41598-025-87782-2Heavy hexagonal codeQuantum circuitsError correctionDeep learningStatistical features |
| spellingShingle | Umesh Uttamrao Shinde Ravikumar Bandaru Re-locative guided search optimized self-sparse attention enabled deep learning decoder for quantum error correction Scientific Reports Heavy hexagonal code Quantum circuits Error correction Deep learning Statistical features |
| title | Re-locative guided search optimized self-sparse attention enabled deep learning decoder for quantum error correction |
| title_full | Re-locative guided search optimized self-sparse attention enabled deep learning decoder for quantum error correction |
| title_fullStr | Re-locative guided search optimized self-sparse attention enabled deep learning decoder for quantum error correction |
| title_full_unstemmed | Re-locative guided search optimized self-sparse attention enabled deep learning decoder for quantum error correction |
| title_short | Re-locative guided search optimized self-sparse attention enabled deep learning decoder for quantum error correction |
| title_sort | re locative guided search optimized self sparse attention enabled deep learning decoder for quantum error correction |
| topic | Heavy hexagonal code Quantum circuits Error correction Deep learning Statistical features |
| url | https://doi.org/10.1038/s41598-025-87782-2 |
| work_keys_str_mv | AT umeshuttamraoshinde relocativeguidedsearchoptimizedselfsparseattentionenableddeeplearningdecoderforquantumerrorcorrection AT ravikumarbandaru relocativeguidedsearchoptimizedselfsparseattentionenableddeeplearningdecoderforquantumerrorcorrection |