The objective of the school is to provide an overview of the main scheduling algorithms and resource management policies for supporting the development of predictable, as well as flexible, real-time systems with hard and soft timing constraints.
The school is aimed at graduate students, PhD students and engineers with some background on computer science and operating systems. No specific knowledge is required to understand the course, since all new concepts are explained and illustrated with concrete examples.
The school is divided in 5 courses, one per day, of 6 hours each:
- Fixed Priority Scheduling
This part introduces the basic concepts of real-time computing systems and provides a theoretical background for analyzing fixed priority systems. Different task models are presented to apply the analysis in different contexts, including resource constraints, release jitter, and fault tolerant requirements. The course also discusses real-time programming issues and advanced research topics.
- Dynamic Priority Scheduling
- Off-line Scheduling
This part focuses on table driven scheduling for time critical applications. It presents the time triggered and the event triggered activiation paradigm, together with a method for constructing an offline schedule based on complex timing constraints. It also presents a method for integrating offline and online scheduling to increase flexibility of static scheduled systems.
- Real-Time Networks
This part introduces the basic concepts of real-time communication networks, focusing on real-time message scheduling, event/time-triggered communication, communication control and synchronization issues. It also presents some common application service models and some case studies of standard communication protocols. It concludes presenting new trends in wireless networks, sensor webs, and internet connection.
- QoS management
This part introduces the problem of resource management for QoS control in centralized systems and distributed systems. Existing solutions are presented for both cases. Middleware architectures for distributed real-time systems are also presented, with particular emphasis on RMI, RT-Java, and RT-RMI.