Detail
Background
In recent years, automotive and aerospace systems have started to use high-performance System-on-Chip platforms and operating systems. These environments often run multiple applications with different levels of criticality at the same time. Such systems are called mixed-criticality systems.
When multiple applications are integrated on one platform, real-time performance and safety become difficult to guarantee. Applications may interfere with one another through shared CPUs, memory, GPUs, storage devices, and network bandwidth.
Detail
Why it is difficult
Interference on shared resources can delay important tasks. In serious cases, this can break timing guarantees, spread faults across applications, or open a path for security attacks.
This means that performance is not only a speed problem. It is also a safety and trust problem, especially in systems that control vehicles and other important equipment.
Detail
What we do
We measure and analyze how real-time applications access shared resources and evaluate which factors reduce predictable behavior. This includes CPUs, memory systems, GPUs, storage, and communication paths.
We also study ways to improve real-time behavior from a full-stack perspective. Topics include software isolation mechanisms, operating system and driver improvements, and hardware support for better timing control.
