White Paper

Whitepaper: Building Safe and Secure Systems for Autonomous Platforms

Issue link: https://read.uberflip.com/i/1427516

Contents of this Issue

Navigation

Page 3 of 7

WHITE PAPER Building Safe and Secure Processing Systems for Future Autonomous Platforms mrcy.com 4 mrcy.com SYSTEM SECURITY System security falls into two domains, cyber (software) and physical. Protecting against remote attacks with cybersecurity Cybersecurity guards against an adversary who attacks a target system through local and network interfaces, which may be wireless. A cyber adversary acts through software tools that originate outside the target system and work their way inward through successive exploitation of system flaws, allowing the adversary multiple levels of trust, permission, access and control. CYBERSECURITY FOCUSES ON SYSTEM INTERFACE PROTECTION THAT INCLUDES: ▪ Verifying that incoming data conforms to the prescribed protocols with errors being properly rejected and/or isolated ▪ Ensuring that capabilities exposed to users align with their approved level of access and authentication ▪ Taking precautions to ensure that any compromise cannot result in an adversary obtaining a persistent system foothold Guarding against attacks with physical security Physical security protects against an adversary who has physical access to the target system. Although a complementary domain to cybersecurity, physical security overlaps cybersecurity as physical possession of a system typically grants the adversary access to all interfaces. In fact, a physical adversary may disassemble a system to gain access to hidden and private interfaces. Intel Security Technologies Enhance Cybersecurity Intel security technologies can help reduce the attack surface area of a computer 's platform stack by providing protection of the boot sequence, data, applications, and workloads System Memory CPU Secure Memory Secure Data and applications Secure Workloads Secure Boot App & Data Core Boot Firmware App & Data Core App & Data Core Cache BEST DO-254/178C DESIGN PROCESS PRACTICES FOR FLIGHT-SAFETY CERTIFICATION INCLUDE: ▪ Traceability between requirements and design implementation, and verification that design intent has been achieved ▪ Justification for including functional elements that do not implement higher-level requirements ▪ Robust testing that characterizes system behavior when confronted with erroneous inputs ▪ Mechanisms that maintain design element configurations for problem tracking and their resolution ▪ Evaluations, tests and decisions that must be meticulously documented and performed with as much organizational independence as possible (e.g., designers do not run tests) Modern processing systems are complex and have so many behavior permutations that it is impossible to test every possible combination. DO-178 and DO-254 are implemented to assure the safety of the aircraft components in the face of these challenges. However, many systems have been operating since before these standards became mandatory. In these cases, if a system has operated as expected across multiple platforms, for long periods, under a range of conditions and dealt with a wide variety of system inputs, the likelihood of a design error is deemed extremely small and its systems inherently safe. Airworthiness authorities may use this product service experience (PSE) approach as criterion for airworthiness. RTCA standards guide design deviation and change management to ensure that lessons learned are fed back and applied throughout the development process. Functional requirements and behavior for each safety-critical element must be documented and validated for completeness and correctness. Consequently, design for certifiable safety is a complex, meticulous and investment-intensive endeavor.

Articles in this issue

Links on this page

view archives of White Paper - Whitepaper: Building Safe and Secure Systems for Autonomous Platforms