智能网联汽车内生安全问题与对策

Endogenous security problems and countermeasures of intelligentconnected vehicle

智能网联汽车正成为新一代典型信息物理系统,面临着传统功能安全与现实网络安全甚至严峻信息安全等二重乃至三重交织问题复合叠加影响,导致这一现象的根本原因之一是智能网联汽车的内生安全问题—软件定义汽车趋势下的软硬件代码设计缺陷、漏洞/后门等问题。在分析智能网联汽车内生安全问题和现有网络安全防御局限性的基础上,提出了“不完全交集”原理和基于动态异构冗余架构的智能网联汽车内生安全理论和方法,通过将网络攻击带来的不确定性失效与软硬件随机性失效归一化为可用概率表达的广义鲁棒控制问题,使安全性指标达到可量化设计和可验证度量的程度,为一体化解决智能网联汽车网络安全和功能安全(信息安全)交织叠加难题提供了新路径,赋能智能网联汽车及车联网基础设施构建数字安全底座。

Intelligent connected vehicle(ICV)is becoming a new generation of typical cyber-physical system,facing the intertwined influence of traditional functional safety,real cybersecurity,and even severe information security problems.One of the roots of this phenomenon is the endogenous security problem of ICVs,namely,problems(e.g.,code defects,vulnerabilities/backdoors,etc.)under the trend of software defined vehicles(SDV).Analyzing the endogenous security problem of ICVs and the limitations of existing network security defense methods,this paper innovatively proposes the incomplete intersection principle(IIP)and the endogenous security theory and method based on dynamic heterogeneous redundant(DHR)architecture.This architecture normalizes both uncertain failures caused by network attacks and random failures of software and hardware into a probabilistic generalized robust control problem,so that the endogenous security indicators are quantifiable and verifiable.Meanwhile,the DHR architecture provides an integrated manner to solve the security and safety intertwine problems(SSIP)of ICVs,enabling ICVs and associated infrastructure to lay the cornerstone of digital security.