https://www.mdu.se/

In this paper, we present a new reservation based scheduling framework for soft real-time systems using EDF algorithm (called CARB-EDF). This framework has the features of Capacity Adaptation, Reclaiming and Borrowing. This framework can simplify the initial configuration of the system, where the system designer does not need to provide any estimations of task execution times. We also present a Chebyshev’s inequality based predictor to estimate task execution times. A number of simulation-based experiments have been implemented. According to the results compared with some related works, our scheduling framework can provide a better performance with acceptable extra scheduling overhead. 

This paper presents a server-based framework for task overrun management in multi-core real-time systems. Unlike most existing scheduling methods which usually assume a single upper bound of the Worst-Case Execution Time (WCET) for each task, our approach targets scenarios with task overruns. The main idea of our framework is to employ Synchronized Deferrable Servers (SDS) to deal with globally scheduled task overruns, while a partitioned scheduling approach is applied on regular task executions. Moreover, we provide a deterministic Worst- Case Response Time (WCRT) analysis focusing on hard timing constraints, along with a probabilistic analysis of Deadline Miss Ratio (DMR) for soft real-time applications. In the evaluation phase, we have implemented two types of experiments evaluating different timing constraints. 

The Controller Area Network (CAN) is widely utilized in many industrial real-time applications. As a real-time communication network, the predictability of timing behaviors is very important. Therefore, many works have been proposed regarding the schedulability analysis of CAN messages. Most of the existing analysis approaches are based on a traditional periodic message model. However, in some applications, the transmission of a message may follow a certain pattern instead of repeating the same transmission period by period. In these cases, applying the existing analysis methods may lead to quite pessimistic results. In order to tackle this problem, in this paper we apply the Generalized Multi-Frame (GMF) task model on CAN messages, where both priority-based and FIFO-based message queues are taken into account. We present a corresponding sufficient schedulability analysis. According to the experimental evaluations, our analysis can provide tighter results compared to the existing CAN message response time analysis in the context of GMF-modeled messages. 

The predictability of timing behavior is a very important performance issue of a real-time system. As the complexity of modern industrial systems increases, analyzing the timing behaviors of those systems becomes more and more challenging. Most of the existing analysis methods depend on static and detailed information of the systems under analysis. However, sometimes only partial information of a system can be available, or it may require too much effort on obtaining those details, making those analysis methods much less feasible. Moreover, those methods usually focus on some specific system models with unrealistic assumptions, consequently, applying those methods on a complex industrial real-time system may result in overly pessimistic results. Therefore, in this paper, we propose a statistical method to compute Worst-Case Response Times (WCRTs) of complex real-time systems regarding soft timing constraints, which can provide a higher general applicability with less required system information. Our approach employs a Peak Over Thresholds (POT) method, which is a branch of the Extreme Value Theory (EVT). For the evaluation, we have applied this approach on the analysis of message transmission latencies over Controller Area Networks (CAN).

In recent years, the complexity of real-time embedded systems has increased dramatically. For those modern real-time systems, the limitations of original static Response Time Analysis (RTA) become more and more conspicuous. Most static analysis methods not only require much detailed system information, but also only target to some specific system model with non-realistic assumptions. As a result, those methods may produce overly pessimistic results, making them unsuitable to be applied on a complex industrial system. The best system model may be the system itself. Therefore, statistical RTA, which can produce probabilistic analysis results based on samples provided by real systems or simulators, may become more expedient. Statistical RTA usually requires more relaxed assumptions and less system information than static RTA. In this paper, we present an Extreme Value Theory (EVT) based method to compute Worst-Case Response Time (WCRT) targeting complex real-time systems. In the evaluation phase, we have applied this method to the calculation of worst-case transmission delays of messages over Controller Area Network (CAN), and some comparisons with static RTA are also provided. According to the experimental results, as the system complexity increases, our approach performs much more stable and less pessimistic.