The world’s first on-chip and in-life monitoring solution to rapidly detect cyber security threats in Connected and Autonomous Vehicles (CAVs)

作为物联网（IoT）的一部分，联网和自动驾驶汽车（cav）代表了一个重大的网络安全挑战。2015年，Twitter安全研究员查理·米勒（Charlie Miller）和IOActive汽车安全研究主管克里斯·瓦拉塞克（Chris Valasek）通过远程入侵菲亚特-克莱斯勒吉普切诺基（Fiat-Chrysler Jeep Cherokee），暴露了汽车的安全漏洞，控制了汽车的各种控制装置，从收音机音量到刹车和方向盘。从那以后，几乎所有的原始设备制造商都遭受了类似的网络攻击。网络安全攻击的社会经济影响是巨大的，据估计，到2023年，汽车行业每年将损失260亿英镑。片上系统（soc）集成了所有计算组件，这些组件通过物联网（IoT）在数字连接的社会框架中为当今的自动驾驶汽车提供动力。对soc“在生活中”（即在操作环境中而不是在实验室中）进行监控的市场需求明显存在，目前的国际标准（ISO 26262“道路车辆-功能安全”）和即将出台的标准（ISO 21434“道路车辆-网络安全工程”）进一步规定了这一点。然而，现有的软件监控解决方案：需要数百毫秒来检测异常，这对于安全关键应用程序来说是不够的。在70英里/小时的速度下，自动驾驶汽车将在700毫秒内行驶22米，这对于避免事故至关重要。影响SoC的正常运行。都很显眼，也更容易被黑客攻击。无法监控整个SoC。目前，还没有基于硬件的实时SoC监控解决方案。UltraSoC提供片上监测解决方案，在设计阶段监测soc的健康状况，我们的客户包括英特尔，海思（华为），C-SKY（阿里巴巴）等大公司。然而，与我们的竞争对手不同，我们的IP是运行时可配置的，供应商中立的，并且基于硬件的允许我们以“时钟速度”（即以微秒为单位）检测异常，这比软件快1000倍。在设计阶段嵌入SoC本身为我们提供了一个独特的机会，利用我们现有的IP来监控SoC的生命健康状况。然而，可靠地监测soc在生活中提出了重大的技术挑战，特别是在不断变化的操作环境中区分正常和异常的系统行为。该项目将开发世界上第一个芯片上和终身监控解决方案，以时钟速度检测系统异常，是供应商中立的，非侵入性的，运行时可配置的，并且不容易被黑客攻击。这将确保物联网系统按其设计功能运行，保护公民和基础设施。

Connected and Autonomous Vehicles (CAVs), as part of the Internet of Things (IoT), represent a major cyber security challenge. In 2015, Charlie Miller, a security researcher at Twitter, and Chris Valasek, director of Vehicle Security Research at IOActive, exposed the security vulnerabilities in automobiles by hacking into a Fiat-Chrysler Jeep Cherokee remotely, controlling the cars' various controls from the radio volume to the brakes and steering wheel. Since then, nearly all OEMs have suffered similar cyber-attacks. The socio-economic impacts of cyber security attacks are tremendous, with the automotive industry estimated to lose £26 billion annually by 2023 [Upstream Security].Systems-on-chip (SoCs) integrate all computing components that power today's CAVs operating within a digitally connected societal framework via the Internet of Things (IoT). A clear market need exists to monitor SoCs 'in-life' (i.e. in an operating environment rather than in the lab), and this is further mandated by current international standards (ISO 26262 "Road Vehicles - Functional Safety") and forthcoming standards (ISO 21434 "Road Vehicles - Cybersecurity engineering").However, existing software monitoring solutions:Take hundreds of milliseconds to detect anomalies, which is inadequate for safety-critical applications. At ~70 mph, a CAV will travel 22 meters in 700 milliseconds, which could be critical to avoid an accident.Disturb the normal operation of the SoC.Are highly visible and more prone to hacking.Are unable to monitor the entire SoC.Today, no hardware-based, in-life SoC monitoring solution exists.UltraSoC provides on-chip monitoring solutions to monitor the health of SoCs during the design phase and our customers include major firms such as Intel, HiSilicon (Huawei), C-SKY (Alibaba) and more. However, unlike our competitors, our IP is runtime configurable, vendor-neutral, and being hardware-based allows us to detect anomalies at 'clock-speed' (i.e. in microseconds), which is 1,000x faster than software.Being embedded into the SoC itself during the design phase gives us a unique opportunity to exploit our existing IP to monitor SoC health in-life. However, reliably monitoring SoCs in-life poses significant technical challenges, especially to discriminate normal and abnormal system behaviour in constantly changing operating environments.This project will develop the world's first on-chip and in-life monitoring solution to detect system anomalies at clock-speed, be vendor-neutral, non-intrusive, runtime configurable and far less prone to hacking. This will ensure IoT systems function as they were designed, protecting citizens and infrastructure.