Citation Context

...recedence constraint, but it does not support fault-tolerance. Dima et al. devised an offiine real-time and fault-tolerant scheduling algorithm using replication of operations and data communications =-=[7]-=-. However, this algorithm must execute the backup copy of a task simultaneously with its primary copy. Manimaran et al. [17] and Mosse et al. [9] have proposed dynamic algorithms to schedule realtime ...

Citation Context

... time between repairs), should determine which subsystems have a compound MTBF that is so short to be of concern. The use of failure patterns to capture effectively these interactions was proposed in =-=[8]-=-, which is the basis of our approach. Setup. Consider the feedback control system in Figure 1. The control system repeats the following sequence at each period Tmax: (1) sensors are sampled, (2) softw...

Citation Context

... which was not the case in [12, 13]. In conclusion, the method presented here is complementary and more general than the one presented in [12, 13]. There is another work involving some of the authors =-=[8]-=-, where a totally different approach is taken: First, communication link failures are also taken into account, and second, the method presented involves building a basic schedule for each possible fai...

Citation Context

...t tolerant deployment of a feedback control system given as a FTDF graph by specifying the fault tolerant binding. A fault tolerant binding guarantees that for each fault scenario, or failure pattern =-=[5]-=-, the execution of a corresponding subset of the actors in FTDF graph, G must be guaranteed. Figure 3 illustrates a FTDF graph for a paradigmatic feedbackcontrol application, the inverted pendulum con...

Citation Context

... and timetriggered protocols [22]. It models mutual exclusion but does not address specific protocol constraints, like time slots reuse or priority-based arbitration. In fact, we build on the work in =-=[23]-=- extending the model in different ways and implementing automation of the methods. First, we introduced FTDF, a MoC that specifically targets real-time control systems where sensors and actuators have...

Citation Context

...ded applications on multicomponent architectures. The method we present extends (Girault et al., 2001; Girault et al., 2003) by tolerating also communication links failures, and is more general than (=-=Dima et al., 2001-=-) since it can tolerate an arbitrary number of processors failures and an arbitrary number of communication links failures. The paper is organized as follows. Section 2 describes the related work. Sec...

Citation Context

...performance implications of allowing CTs and NCTs on the same resources in a fault-tolerant system. Several reliability techniques use sets of static schedules to dynamically respond to failure [10], =-=[13]-=-, [14]. Another approach adjusts a static schedule at runtime to enhance reliability, allowing tasks to re-execute by moving the execution of other tasks [15]. Alternatively, our approach performs sta...

Citation Context

...tion of critical and non-critical tasks, reducing the overhead of redundancy by increasing NCT execution. Several reliability techniques use sets of static schedules to dynamically respond to failure =-=[4, 16, 19, 20]-=-. Another approach adjusts a static schedule at runtime to enhance reliability, allowing tasks to re-execute by moving the execution of other tasks [21]. More recent research has investigated employin...