Cyber Security Analytics: Human Vulnerability to Cyber Attack Attempts When Using Autonomous Vehicles

网络安全分析：使用自动驾驶汽车时人类对网络攻击尝试的脆弱性

• 批准号: 2598912
• 金额: --
• 依托单位: CARDIFF UNIVERSITY
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2021
• 资助国家: 英国
• 起止时间: 2021 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2598912
• 关键词: Cyber; Security; Analytics; Human; Vulnerability

Autonomous vehicles (AVs) promise huge benefits from improved road safety to reduced carbon emission. To reap the full benefit of such technology, future AVs have been envisioned to be "connected" in the sense that they will be able to communicate with each other and the external infrastructure through shared wireless networks so that optimisation of collective decision making can be achieved. This introduces new threats of cyber-attacks which could lead to severe disruptions to system operation and potentially cause catastrophic incidents. The most common infiltration/attack techniques (e.g., propagating malware across the entire network using spamming, malicious messaging and update requests) usually rely on the "cooperation" of the user of a terminal computer (e.g., accepting a request), which makes the human potentially the weakest link of the system if they succumb to such malicious attempts with data showing human error can be responsible for up to 95% of cyber security breaches. Human-factor-dependent cyber security vulnerabilities can be exacerbated in the context of autonomous driving - the low cognitive demand from the driving task may cause a lowered situation awareness and trigger an automatic, heuristic-driven cognitive processing style, which could make AV users more susceptible to falling victim of malicious spamming, as research shows.

自动驾驶汽车（AV）承诺从改善道路安全到减少碳排放的巨大好处。为了获得这种技术的全部好处，未来的AV已经被设想为“连接”，在这个意义上，它们将能够通过共享的无线网络与彼此和外部基础设施进行通信，从而可以实现集体决策的优化。这带来了新的网络攻击威胁，可能导致系统运行严重中断，并可能导致灾难性事件。最常见的渗透/攻击技术（例如，使用垃圾邮件、恶意消息和更新请求在整个网络上传播恶意软件）通常依赖于终端计算机用户的“合作”（例如，接受请求），这使得人类可能是系统中最薄弱的环节，如果他们屈服于这种恶意尝试，数据显示人为错误可能导致高达95%的网络安全漏洞。在自动驾驶的背景下，依赖于人为因素的网络安全漏洞可能会加剧-驾驶任务的低认知需求可能会导致情境意识降低，并触发自动的，主动驱动的认知处理方式，这可能使AV用户更容易成为恶意垃圾邮件的受害者，正如研究所示。