|
|
大模型在嵌入式软件开发中的应用
|
Abstract:
嵌入式软件与硬件和领域知识高度相关，开发难度大，开发周期长。以Transformer架构为核心的大语言模型(LLMs)凭借强大的代码理解与生成能力，为提升嵌入式软件开发效率提供了可能。文章介绍大语言模型在嵌入式软件开发中的应用现状，包括应用过程、方法和常用的大语言模型及其技术特性，并重点梳理大模型在嵌入式软件开发中的应用场景，以及嵌入式软件代码缺陷检测的核心技术路径、场景适配方案。文章通过综述现有研究成果，为嵌入式软件研究和开发人员应用大模型提供参考，助力嵌入式软件开发方法的智能化转型。
Embedded software is highly correlated with hardware and domain knowledge, featuring high development difficulty and long development cycles. Large Language Models (LLMs) centered on the Transformer architecture, with their robust capabilities in code understanding and generation, have made it possible to improve the efficiency of embedded software development. This paper introduces the current application status of large language models in embedded software development, including the application processes, methods, commonly used large language models and their technical characteristics. It also focuses on sorting out the application scenarios of large models in embedded software development, as well as the core technical approaches and scenario adaptation schemes for code defect detection in embedded software. Through a review of existing research findings, this paper provides a reference for embedded software researchers and developers in the application of large models, and facilitates the intelligent transformation of embedded software development methodologies.
| [1] | Jiang, J., Wang, F., Shen, J., et al. (2024) A Survey on Large Language Models for Code Generation. https://dl.acm.org/doi/full/10.1145/3747588 |
| [2] | Hui, B., Yang, J., Cui, Z., et al. (2024) Qwen2.5-Coder Technical Report. https://arxiv.org/abs/2409.12186 |
| [3] | Englhardt, Z., Li, R., Nissanka, D., Zhang, Z., Narayanswamy, G., Breda, J., et al. (2024) Exploring and Characterizing Large Language Models for Embedded System Development and Debugging. Extended Abstracts of the CHI Conference on Human Factors in Computing Systems, Honolulu, 11-16 May 2024, 1-9. https://doi.org/10.1145/3613905.3650764 |
| [4] | Stop Wasting Hours Debugging: Use AI to Write Better Embedded Systems Code. https://markaicode.com/ai-embedded-systems-code/ |
| [5] | Xu, R., Cao, J., Wu, M., et al. (2025) Embed Agent: Benchmarking Large Language Models in Embedded System Development. https://arxiv.org/abs/2506.11003 |
| [6] | Patil, M.S., Ung, G. and Nyberg, M. (2024) Towards Specification-Driven LLM-Based Generation of Embedded Automotive Software. In: Lecture Notes in Computer Science, Springer, 125-144. https://doi.org/10.1007/978-3-031-75434-0_9 |
| [7] | Medaranga, P., Shah, D., Kandala, S.V. and Varshney, A. (2024) POSTER: Simplifying the Networking of Wireless Embedded Systems Using a Large Language Model. Proceedings of the ACM SIGCOMM 2024 Conference: Posters and Demos, Sydney, 4-8 August 2024, 78-80. https://doi.org/10.1145/3672202.3673752 |
| [8] | Lin, J., Luo, Y., Chen, X., Gu, B. and Jin, Z. (2025) Automatic Generation of Structured Requirements for Aerospace Embedded Systems Using LLMs. 2025 IEEE 33rd International Requirements Engineering Conference Workshops (REW), Valencia, 1-5 September 2025, 181-188. https://doi.org/10.1109/rew66121.2025.00028 |
| [9] | Yang, H., Li, M., Han, M., et al. (2024) Embed Genius: Towards Automated Software Development for Generic Embedded IoT Systems. https://arxiv.org/abs/2412.09058 |
| [10] | MITRE Corporation and SANS Institute (2024) 2024 CWE Top 25 Most Dangerous Software Weaknesses. https://cwe.mitre.org/top25/archive/2024/2024_cwe_top25.html |
| [11] | National Institute of Standards and Technology (2022) Security and Privacy Controls for Information Systems and Organizations (NIST SP 800-53 Revision 5). https://csrc.nist.gov/publications/detail/sp/800-53/rev-5/final |
| [12] | Bhandari, G.P., Assres, G., Gavric, N., Shalaginov, A. and Grønli, T. (2024) IoTvulCode: AI-Enabled Vulnerability Detection in Software Products Designed for IoT Applications. International Journal of Information Security, 23, 2677-2690. https://doi.org/10.1007/s10207-024-00848-6 |
| [13] | Yang, Y.L. (2023) IoT Software Vulnerability Detection Techniques through Large Language Model. In: Lecture Notes in Computer Science, Springer, 285-290. https://doi.org/10.1007/978-981-99-7584-6_21 |
| [14] | Hanif, H. and Maffeis, S. (2022) VulBERTa: Simplified Source Code Pre-Training for Vulnerability Detection. 2022 International Joint Conference on Neural Networks (IJCNN), Padua, 18-23 July 2022, 1-8. https://doi.org/10.1109/ijcnn55064.2022.9892280 |
| [15] | Qayyum, K., Jha, C.K., Ahmadi-Pour, S., Hassan, M. and Drechsler, R. (2025) LLM-Assisted Bug Identification and Correction for Verilog HDL. ACM Transactions on Design Automation of Electronic Systems, 30, 1-28. https://doi.org/10.1145/3733237 |
| [16] | Qayyum, K., Hassan, M., Ahmadi-Pour, S., Jha, C.K. and Drechsler, R. (2024) From Bugs to Fixes: HDL Bug Identification and Patching Using LLMs and Rag. 2024 IEEE LLM Aided Design Workshop (LAD), San Jose, 28-29 June 2024, 1-5. https://doi.org/10.1109/lad62341.2024.10691874 |
| [17] | Cui, F., Yin, C., Zhou, K., Xiao, Y., Sun, G., Xu, Q., et al. (2024) OriGen: Enhancing RTL Code Generation with Code-to-Code Augmentation and Self-Reflection. Proceedings of the 43rd IEEE/ACM International Conference on Computer-Aided Design, New York, 27-31 October 2024, 1-9. https://doi.org/10.1145/3676536.3676830 |
| [18] | Tsai, Y.D., Liu, M. and Ren, H. (2024) RTLFixer: Automatically Fixing RTL Syntax Errors with Large Language Model. Proceedings of the 61st ACM/IEEE Design Automation Conference, San Francisco, 23-27 June 2024, 1-6. https://doi.org/10.1145/3649329.3657353 |
| [19] | Thakur, S., Blocklove, J., Pearce, H., et al. (2023) AutoChip: Automating HDL Generation Using LLM Feedback. https://arxiv.org/abs/2311.04887 |
