Research On Open Real-time System Schedule Framework And Algorithms

Real-time systems are mainly designed to satisfy the timing requirements from the real world applications. They are widely adopted in many areas, such as industrial controls, aerospace electronics, military equipments, and so on. With the development of real-time system applications, the cases that different kinds of hard real-time, soft real-time and non-real-time applications coexist in one system become more and more popular. This makes the applications requirements become more complex, then open hybrid real-time system has been proposed.In this thesis, we study the scheduling problems for open hybrid real-time systems, and the main contributions of this dissertation are as follows:1. A new integrated scheduling framework (ORTSSF) has been presented in this paper. In this frame, each real-time task is served by a dedicated server, and all non-real-time tasks are reserved some bandwidth, so all tasks can coexist in a system. And two levels of controller that controls tasks' in or out. The former guarantees that tasks are independent and system schedulability.2. Toward hybrid tasks schedule, we have developed a new algorithm HBASHA, which enhances the Constant Bandwidth Server (CBS) by slack reclaiming. The algorithm is able to guarantee that all hard real-time jobs be completed before their deadlines, improve QoS of soft real-time tasks as much as possible and make non-real-time tasks no starvation.3. In real world, tasks are usually interrelated, and there is resource sharing between them. In this paper, we have implemented SRP in HBASHA, so this algorithm can solve resource sharing problem.4. We have implemented this framework and relative algorithms in RTSim (a real-time system simulator). And we have done experiments to verify the framework.