Energy harvesting is an emerging technology that enhances the lifetime of Internet-of- Things (loT) applications. Satisfying real-time requirements for these systems is challenging. Dedicated real-time schedulers integrating both timing and energy constraints are required, such as the ED- H scheduling algorithm[l]. However, this algorithm has been proved to be optimal for independent tasks only (i.e., without considering any shared resources), thus preventing its confident deployment into computing infrastructures in which tasks are mostly interdependent. In this paper, we first derive worst-case blocking times and worst-case blocking energy for tasks sharing resources managed by the DPCP protocol[2] and scheduled under the ED- H scheme. Then, we provide a sufficient schedulability test for ED-H-DPCP guaranteeing off-line that both timing and energy constraints will be satisfied, even in the presence of shared resources.

Abstract Dynamic voltage and frequency scaling (DVFS) is a promising and broadly used energy-efficient technique to overcome the main problems arising from using a finite energy reservoir capacity and uncertain energy source in real-time embedded systems. This work investigates an energy management scheme for real-time task scheduling in variable voltage processors located in sensor nodes and powered by ambient energy sources. We use DVFS technique to decrease the energy consumption of sensors at the time when the energy sources are limited. In particular, we develop and prove an optimal real-time scheduling framework with speed stretching, namely energy guarantee DVFS (EG-DVFS), that jointly accounts not only for the timing constraints, but also for the energy state incurred by the properties of the system components. EG-DVFS relies on the well-known earliest deadline-harvesting scheduling algorithm combined with DVFS technique where the sensor processing frequency is fine tuned to further minimize energy consumption and to achieve an energy autonomy of the system. Further, an exact feasibility test for a set of periodic, aperiodic or even sporadic tasks is presented.

Abstract We consider the problem of real-time scheduling in uniprocessor devices powered by energy harvesters. In particular, we focus on mixed sets of tasks with time and energy constraints: hard deadline periodic tasks and soft aperiodic tasks without deadlines. We present an optimal aperiodic servicing algorithm that minimizes the response times of aperiodic tasks without compromising the schedulability of hard deadline periodic tasks. The server, called Slack Stealing with energy Preserving (SSP), is designed based on a slack stealing mechanism that profits whenever possible from available spare processing time and energy. We analytically establish the optimality of SSP. Our simulation results validate our theoretical analysis.

Ambient energy harvesting has become a popular solution for battery-operated systems with finite energy supply such as wireless sensor networks. This paper investigates the problem of multiprocessor real-time scheduling in a system whose energy reservoir is replenished by an ambient energy source. In particular, we focus on energy-efficient partitioning for periodic real-time tasks in a homogeneous multi-core platform where both timing and energy requirements are considered. We assume that the optimal scheduler, namely the Earliest Deadline - Harvesting (ED-H) [1], is used on every core of the architecture. Our objective is to find a feasible partitioning solution based on the real-time execution of tasks allocated to the sensor nodes based on the actual energy harvesting data to guarantee the desirable absence of both energy starvation situations and deadline violation. For this sake, we propose an Energy Harvesting-Reasonable Allocation (EH-RA) algorithm that amounts to the traditional bin-packing technique by guaranteeing both timing constraints and energy awareness perspectives. Experimental results show that our approach can achieve a significant gain in performance when compared to EDF.

This paper explores an energy harvesting-based approach for a jointly execution of hard periodic tasks mixed with occurring soft aperiodic tasks in battery-powered embedded devices. Providing an optimal solution to this scheduling issue is challenging. Accordingly, we propose a new aperiodic task scheduling algorithm in real-time energy harvesting (RTEH) systems that targets at minimizing the response times of aperiodic tasks without compromising the schedulability of periodic tasks. Based on the slack stealing mechanism, the proposed algorithm called Slack Stealing with energy Preserving (SSP) profits whenever possible from available extra processing time and available extra energy so as to service the aperiodic tasks as soon as possible. Comparing to other algorithms, SSP achieves better performance in terms of the aperiodic responsiveness under various energy profiles and settings.

This paper explores the energy-aware strategy for adaptive task allocation of sporadic task model in distributed reconfigurable embedded systems powered by a renewable energy source. It is challenging to ensure hard real-time deadlines of sporadic tasks in distributed reconfigurable embedded systems that entirely rely on energy harvesting with limited capacity storage because their release times and periods are unknown beforehand. In this regard, we propose a new scheduling algorithm for sporadic task model in energy harvesting reconfigurable distributed embedded system. The proposed approach generates an energy efficient offline task assignment heuristic. Then, the offline stage is dynamically extended online by applying a dynamic adaption algorithm. The latter performs three solutions: (i) Migration, (ii) Degradation of the execution mode, and (iii) Removal of tasks in order to maintain schedulability and sustainability of the application software. Simulations show the efficiency of the proposed approach in term of deadline success ratio.

In this paper, we are interested in real-time scheduling of a distributed reconfigurable embedded system powered by a renewable energy source. Uncertainty of energy availability in energy harvesting systems makes the problem of task scheduling more challenging. A reconfiguration scenario is defined as an operation that allows the addition-removalmodification of tasks which may result in timing infeasibility. The system updates its behaviour in accordance with user requirements in reaction to unpredictable events from the environment. The contribution of this paper concerns energyaware strategies for adaptive task allocation. The proposed approach is decomposed into two phases a) Static task assignment heuristic, and b) Dynamic adjustment algorithm. In order to maintain schedulability and sustainability of the application software, the dynamic adjustment algorithm performs three functions: i) Migration, ii) Degradation of the execution mode, and iii) Removal of tasks. We employ a novel energy prediction algorithm. A simulation study brings to light the effectiveness of the proposed approach in terms of deadline success ratio.

In this paper, we propose task partitioning heuristics for multicore real-time energy harvesting systems. Both timing constraints and energy requirements are considered. Our objective is to determine a partitioning that guarantees absence of both energy starvation and deadline missing under Earliest Deadline First (EDF) scheduling. First, we describe the system model and formalize the assignment problem in real-time energy harvesting systems. Then, we describe and analyze the performance of partitioning heuristics for identical processors. We explore how the task sorting criteria and the energy constraints can favor the heuristics feasibility performance. Moreover, we estimate through simulation a sufficient energy reservoir capacity that guarantees the best performance.

Energy harvesting is the capture of ambient energy, its conversion into a usable form, and its storage for immediate or future use. Interest in energy harvesting has increased over the last decade because of its environmental friendliness and its ability to power devices without electric wires. This term has been frequently applied in recent years in the context of small autonomous embedded devices such as wireless sensor nodes. In this paper, we address the scheduling problem for a single processor device that executes preemptable time critical tasks. Each one has a certain energy requirement and arrives at an unpredictable time. We ask the question whether the traditional task scheduling algorithm earliest deadline first (EDF) is convenient for energy harvesting environments. The paper shows that EDF has a zero competitive factor but nevertheless is optimal for online non-idling settings.

The need for supporting dynamic real-time environments where changes in workloads may occur requires a scheduling framework that explicitly addresses overload conditions, allows the system to achieve graceful degradation and supports a mechanism capable of determining the load to be shed from the system to handle the overload. In applications ranging from video reception to air-craft control, tasks enter periodically and have response time constraints, but missing a deadline is acceptable, provided most deadlines are met. Such tasks are said to be occasionally skippable and have an assigned skip parameter. We look at the problem of uniprocessor scheduling of skippable periodic tasks which consists in maximizing the robustness of the system defined as the global completion ratio. In this paper, we propose a novel scheduling Skip-over algorithm, called RLP/T, a variant of Earliest-Deadline First which adjusts the system workload such that tasks adhere to their timing and skip constraints and guarantees the best robustness.

Summary form only given. In the last years, multimedia and feedback control applications have emphasized the importance of appropriate resource allocation policies in real-time systems, introducing the problem of controlling and adapting the quality of service (QoS) provided by an application. The basic idea is then to avoid overload situations by scaling down the applications resource requirements, while sustaining a specified QoS level for the system. In this paper, we are interested in the problem of dynamic QoS driven CPU allocation for hybrid sets of tasks, consisting of firm periodic tasks, i.e, tasks allowing occasional skips of instances) and soft aperiodic requests. The approach seeks to establish a compromise between minimizing the response time of aperiodic requests and maximizing the QoS of periodic tasks. First, we briefly present the library of free software components developed within the French National project CLEOPATRE (work supported by the French research office, grant number 01K0742). Then, we focus on the functioning of the QoS management component. Finally, we present some simulation results to underline the interest of such a scheduling scheme for real-time applications' developers.

In this paper, we study the problem of scheduling periodic task sets defined under quality of service (QoS) constraints. In our approach, periodic tasks allow occasional skips of instances. A new algorithm, called RLP (red tasks as late as possible) based on the skip-over model and the EDL (earliest deadline as late as possible) scheduling strategy, is proposed to enhance the QoS observed for periodic tasks, i.e, the ratio of periodic tasks which complete before their deadline. We prove that our results are never worse than those obtained in previous work. Experimental results also show significant improvement achieved by our algorithm over RTO and BWP.

Abmw&-Task scheduling is an important issue in the design of a renl-timc computer system because tasks have execution deadlines that must be met, otherwise the system fails with severe consequences upon the environment. In this paper, we study the problem of scheduling periodic time critical tasks on a monoprocessor system. A periodic time critkal task consists of an infinite number of -quests, each of whieh has a prescribed deadline. Tasks are assumed to meet their timing requirements when scheduled by the Earliest Deadline algorithm and preemptions are allowed. We report results from some investigations into the problem of making optimum use of the remaining processor idle time in scheduling perlodk tasks either as soon as possible M as late as possible. The major results consist of the statement and proof of properties relating to bcdhtion and duration of idle time intervals and enable us to provide an elRcient algorlthm lor determining maximum quantity of total idle time available between any two instants. We describe how these results can be applied, Brst to the decision problem that arises when a sporadic time critical task occurs and requires to be run at an unpredictable time and second, to the scheduling problem that arises in a fault tolerant system using the deadline mechanism for which each task implements primary and alternate algorithms. Index Terms-Deadline mechanism, idle time, preemptive schedul