Scholarlink

ns-3 is a widely used simulator for evaluating novel network protocols and algorithms in both academic and industrial settings, eliminating the need for real-world deployment. Due to its open-source nature and active researcher contributions, new versions are frequently released. While supplementary documentation is provided, manually tracking classes changes across versions remains a significant challenge. To address this, we propose and implement a version compatibility checking system powered by a Large Language Model (LLM). This system utilizes a Retrieval-Augmented Generation (RAG) framework, where a vector database is constructed from the release note of a specific ns-3 version. Furthermore, we design a specialized prompt to analyze submitted source code and determine if its use of ns-3 classes and functions adheres to the requirements of the specified target version.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Leveraging Large Language Models for Automated Source Code Version Checking in ns-3

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • YongBeom Park,
  • HeeWon Choe,
  • SuIn Park,
  • Ki-Il Kim

摘要

ns-3 is a widely used simulator for evaluating novel network protocols and algorithms in both academic and industrial settings, eliminating the need for real-world deployment. Due to its open-source nature and active researcher contributions, new versions are frequently released. While supplementary documentation is provided, manually tracking classes changes across versions remains a significant challenge. To address this, we propose and implement a version compatibility checking system powered by a Large Language Model (LLM). This system utilizes a Retrieval-Augmented Generation (RAG) framework, where a vector database is constructed from the release note of a specific ns-3 version. Furthermore, we design a specialized prompt to analyze submitted source code and determine if its use of ns-3 classes and functions adheres to the requirements of the specified target version.