《计算机应用》唯一官方网站

• •    下一篇

基于区块链的自动驾驶仿真测试数据存证模型

彭海洋1,计卫星2,刘法旺3   

  1. 1. 北京理工大学计算机学院
    2. 北京师范大学人工智能学院
    3. 工业和信息化部 装备工业发展中心
  • 收稿日期:2024-09-09 修回日期:2024-10-24 发布日期:2024-11-19 出版日期:2024-11-19
  • 通讯作者: 彭海洋
  • 基金资助:
    新一代人工智能国家科技重大专项

Blockchain-based model for notarization of simulation testing data in autonomous driving

  • Received:2024-09-09 Revised:2024-10-24 Online:2024-11-19 Published:2024-11-19
  • Supported by:
    National Science and Technology Major Project

摘要: 针对自动驾驶汽车仿真测试中多参与者之间数据共享带来的安全问题,提出了一种基于区块链的自动驾驶仿真测试数据存证模型,利用区块链确保数据的安全存储和可追溯性,为审计工作提供可靠支持。首先,利用联盟链的半公开特性确保仿真测试数据仅对授权组织可见,并通过基于RBAC模型的权限校验机制实现数据的安全访问控制;其次,通过定义链上数据访问的智能合约模板规范了数据存取流程,并通过开放流程扩展点支持自定义功能,针对仿真资源的交易行为,通过关联合约确保交易行为的自动执行;最后,提出了IPFS链上链下混合存储、数据批量处理和资源数据缓存等优化策略解决区块链存储资源和处理能力的限制的问题。实验结果表明,针对500个模拟场景的数据存证过程,优化措施使得总事务量减少了72%,链上数据写入和读取耗时分别降低了85.31%和52.67%

关键词: 自动驾驶仿真测试, 区块链, 智能合约, 数据存证, 数据安全

Abstract: To address the security issues arising from data sharing among multiple participants in autonomous vehicle simulation testing, a blockchain-based data notarization model for autonomous driving simulation testing is proposed. This model leverages blockchain to ensure secure storage and traceability of data, providing reliable support for auditing processes. Firstly, the semi-public nature of consortium blockchain ensures that simulation test data is only accessible to authorized organizations, and a Role-Based Access Control (RBAC) mechanism is implemented to achieve secure data access control. Secondly, the data access process is standardized through the definition of smart contract templates on the blockchain, and custom functionalities are supported via open process extension points. For the trading of simulation resources, associated contracts are employed to ensure the automatic execution of transactions, facilitating seamless resource access within the simulation process. Lastly, the optimization strategies, including IPFS-based hybrid on-chain and off-chain storage, batch data processing, and resource data caching, address the limitations of blockchain storage resources and processing capabilities. Experimental results demonstrate that for the notarization of data in 500 simulated scenarios, these optimization measures lead to a 72% reduction in transaction volume, with the time required for data writing and reading on the blockchain reduced by 85.31% and 52.67%, respectively.

Key words: autonomous driving simulation testing, blockchain, smart contracts, data notarization, data security

中图分类号: