Real-Time Embedded Systems Concepts and Practices

This introductory module covers fundamental real-time theory and challenges in designing real-time systems. Topics include predictable response software, comparison of Linux POSIX real-time threads to RTOS and MFE systems, scheduling theory for real-time systems, Rate Monotonic optimal priority policy, service request and response timeline analysis, and challenges with Linux user and kernel space real-time applications. Students gain practical experience through programming assignments focused on creating and managing threads in Linux environments.

This module explores quality of service principles for real-time systems, focusing on utility curves, Rate Monotonic Scheduling Policy, and the Least Upper Bound Condition. Topics include code demonstration comparing SMP to AMP architectures, thread affinity implementation, real-time clock usage, absolute time and date standards, and Rate Monotonic Scheduling analysis methods. Students learn to evaluate system feasibility and safety, particularly when exceeding the Rate Monotonic Least Upper Bound, through theoretical study and practical programming assignments.

This module focuses on scheduling methods and real-time service implementation. Topics include priority preemptive scheduler state machines for Linux and VxWorks, interrupt-driven service code structures, Rate Monotonic Analysis limitations and workarounds, POSIX software interval timers, generic sequencing techniques, comparison of cyclic executive implementations, worst-case analysis through hand diagramming, and introduction to the Cheddar RMA analysis tool. Students learn to analyze various scheduling scenarios and implement real-time services.

This final module examines hardware platforms and software environments for real-time applications. Topics include ARM processor families (M-Series, R-Series, and A-Series), single-board computers for real-time education, Linux usage rationale, and thread gridding concepts for multi-core systems. The module also covers the Space Shuttle Multi-Frequency Executive Architecture as a case study. Students complete a programming assignment implementing CPU core affinity in Linux to emulate Asymmetric Multi-Processing, gaining practical experience with thread distribution across multiple cores.