Search

Targeting global EDF scheduling, this paper proposes an optimal algorithm for parallelizing tasks with parallelization freedom. For this, we extend the interference-based sufficient schedulability analysis and derive monotonic increasing properties of both tolerance and interference for the schedulability. Leveraging those properties, we propose a one-way search based optimal algorithm with polynomial time complexity. We present a formal proof of the optimality of the proposed algorithm. We first address the single-segment task model and then extend to the multi-segment task model. Our extensive experiments through both simulation and actual implementation show that our proposed approach can significantly improve the schedulability.

Due to the complex computation requirements of autonomous car functions, high-performance ECUs are inevitable for the future automotive system architecture. However, it has been reported that high-performance ECUs are usually more susceptible to the soft-error due to the technology scaling. One of the key methods to tolerate these soft errors (i.e. transient faults) is the lockstep mechanism based on the hardware redundancy. However, since this uses multiple redundant cores to execute a single program, it is very inefficient in terms of resource usage. To cope with this drawback, a new feature, ‘Lockstep/Performance mode switch' was introduced. It aims to guarantee safety for safety-critical tasks using the lockstep and maximize throughput for non-critical tasks using different cores for different tasks. This paper proposes a real-time scheduling method in order to efficiently use this new feature following the previous work. We first propose a condition of Periodic Resource Model(PRM) that avoids pessimism while we construct a PRM with the harmonic task set. Then, we partition the given input set into multiple PRMs where each PRM possesses a harmonic task set and has the optimal PRM period and budget. Consequently, The proposed method achieves optimal maximum schedulable utilization, which is more resource-efficient than the previous work.

Recent parallel programming frameworks such as OpenCL and OpenMP allow us to enjoy the parallelization freedom for real-time tasks. The parallelization freedom creates the time versus density tradeoff problem in fluid scheduling, i.e., more parallelization reduces thread execution times but increases the density. By system-widely exercising this tradeoff, we propose optimal parameter tuning of real-time tasks aiming at maximizing the schedulability of multicore fluid scheduling. Our experimental study by both simulation and actual implementation shows that the proposed approach well balances the time and the density, and results in up to 80 percent improvement of the schedulability.

Targeting global EDF scheduling, this paper proposes a conditionally optimal algorithm for parallelizing tasks with parallelization freedom. For this, we extend the interference-based sufficient schedulability analysis and derive monotonic increasing properties of both tolerance and interference for the schedulability. Leveraging those properties, we propose a one-way search based conditionally optimal algorithm with polynomial time complexity. Our extensive experiments through both simulation and actual implementation show that our proposed approach can significantly improve the schedulability up to 60 percent.

IEEE 802.15.4 is a global standard designed for emerging applications in low-rate wireless personal area networks (LRWPANs). The standard provides beneficial features, such as a beacon-enabled mode and guaranteed time slots for realtime data delivery. However, how to optimally operate those features is still an open issue. For the optimal operation of the features, this paper proposes a holistic optimization method that jointly optimizes three cross-related problems: cluster-tree construction, nodes’ power configuration, and duty-cycle scheduling. Our holistic optimization method provides a solution for those problems so that all the real-time packets can be delivered within their deadlines in the most energyefficient way. Our simulation study shows that compared to existing methods, our holistic optimization can guarantee the on-time delivery of all real-time packets while significantly saving energy, consequently, significantly increasing the lifetime of the network. Furthermore, we show that our holistic optimization can be extended to take advantage of the spatial reuse of a radio frequency resource among long distance nodes and, hence, significantly increase the entire network capacity. Category: Ubiquitous computing

The current simulation tools used in the automotive industry do not correctly model timing behaviors of cyber-systems such as varying execution times and preemptions. Thus, they cannot correctly predict the real control performance. Motivated by this limitation, this paper proposes functionally and temporally correct simulation for the cyber-side of an automotive system. The key idea is to keep the data and time correctness only at physical interaction points and enjoy freedom of scheduling simulated jobs for all other cases. This way, the proposed approach significantly improves the real-time simulation capacity of the state-of-the-art simulation methods while keeping the functional and temporal correctness.

This paper proposes a novel technique called mRT-PLRU (Multitasking Real-Time constrained combination of Pinning and LRU), which forms a generic framework to use inexpensive nonvolatile NAND flash memory for storing and executing real-time programs in multitasking environments. In order to execute multiple real-time tasks stored in NAND flash memory with the minimal usage of expensive RAM, the mRT-PLRU is optimally configured in two steps. In the first step, the per-task analysis finds the function of RAM size versus execution time (and the corresponding optimal pinning/LRU combination) for each individual task. Using these functions for all the tasks as inputs, the second-step called a stochastic-analysis-in-loop optimization conducts an iterative convex optimization with the stochastic analysis for the probabilistic schedulability check. As a result, the optimization loop can optimally determine the RAM sizes for multiple tasks such that their deadlines are probabilistically guaranteed with the minimal size of total RAM. The usefulness of the developed technique is intensively verified through both simulation and actual implementation. Our experimental study shows that mRT-PLRU can save up to 80 percent of RAM required by the industry-common shadowing approach.

Wireless sensor networks have great potentials to be applied to mission critical real-time applications such as target tracking, emergency detection, and disaster management. For such real-time applications, sensory data should be delivered within the end-to-end deadlines; otherwise, the data may become useless or misleading. In order to provide such deterministic guarantee of end-to-end deadlines, not only packet scheduling within a single hop but also routing across multiple hops have to be jointly considered. In this paper, we propose a combined scheduling and routing mechanism that provides a deterministic guarantee of end-to-end deadlines leveraging real-time scheduling and analysis techniques. Our experimental study shows that the proposed mechanism can accommodate a larger workload meeting all the end-to-end deadlines than a pretty comprehensive set of other possible solutions of scheduling and routing that are known to deterministically guarantee end-to-end deadlines.

This paper proposes a safe stochastic analysis for fixed-priority scheduling, which is applicable to a broader spectrum of periodic tasks than the ones analyzable by any of the existing techniques. The proposed analysis can find a safe upper-bound of deadline miss probability for periodic tasks with (1) arbitrary execution time distributions, (2) varying interrelease times with the period as the minimum, and (3) the maximum utilization factor Umax that can be greater than 1. One challenge for this is that the release times of tasks are not known a priori because we are not limiting the interrelease times of each task to a constant, i.e., the period. In such a situation, the relative phases of task instances at run time can be arbitrary. Thus, we need to consider all possible phase combinations among jobs to find the worst case deadline miss probability, which is not tractable. To handle this difficulty, we first derive the worst case phase combination for harmonic task sets. Then, we present a safe way to transform a nonharmonic task set to a harmonic task set such that the deadline miss probabilities obtained with the worst case phase combination for the transformed harmonic task set are guaranteed to be worse than those for the original nonharmonic task set with all possible phase combinations. Therefore, the worst case deadline miss probabilities of the transformed harmonic tasks can be used as safe upper-bounds of deadline miss probabilities of the original nonharmonic tasks. Through experiments, we show that the safe upper-bound computed by the proposed analysis is tight enough for practical uses.

As the elderly population increases, elderly care using inexpensive technological means is becoming critical. This paper presents our prototype system that provides real-time indoor tracking of elderly residents and their belongings, which is essential to assisting and securing their independent living. For high-fidelity real-time tracking, we propose novel scheduling algorithms. Our scheduling algorithms are designed by harmonizing both sensing and communication signals and leveraging location awareness and mobility consciousness in order to improve tracking accuracy while reducing the energy consumption. We performed extensive experiments through both simulation and actual implementation. Our experimental result says that our scheduling algorithms can provide real-time tracking of residents within a 20 cm error bound in the typical range of human mobility.

Recent computing applications such as videoconferencing and grid computing run their tasks on distributed computing resources connected through networks. For such applications, knowledge of the network status such as delay, jitter, and available bandwidth can help them select proper network resources to meet the Quality-of-Service (QoS) requirements. Also, the applications can dynamically change the resource selection if the current selection is found to experience poor performance. For such purposes, Internet Service Providers (ISPs) have started to instrument their networks with Network Measurement Infrastructures (NMIs) that run active measurement tasks periodically and/or on demand. However, one problem that most network engineers have overlooked is the measurement conflict problem, which happens when multiple active measurement tasks inject probing packets into the same network segment at the same time, resulting in misleading reports of network performance due to their combined effects. This paper proposes enhanced Earliest Deadline First (EDF) algorithms that allow "Concurrent Executions" to orchestrate offline/online measurement jobs in a conflict-free manner. The simulation study shows that our measurement scheduling mechanism can improve the schedulable utilization of offline measurement tasks up to 300 percent and the response time of on-demand jobs up to 50 percent. Further, we implement and deploy our scheduling mechanism in a real working NMI for monitoring the Internet2 Abilene network. As a case study, we show the utility of our algorithms in the widely used Network Weather Service (NWS).

This paper describes a stochastic analysis framework which computes the response time distribution and the deadline miss probability of individual tasks, even for systems with a maximum utilization greater than one. The framework is uniformly applied to fixed-priority and dynamic-priority systems and can handle, tasks with arbitrary relative deadlines and execution time distributions.

In many practical real-time applications, there is a given set of task frequencies (i.e., inverse of task periods) corresponding to predetermined QoS options that the applications can choose. For example, in audio applications, the typical choices of playback frequencies are for CD quality, radio quality, and phone quality. Similar configurations can be found in streaming video and in control applications. In such systems, applications dynamically arrive requesting one of the periods provided by the QoS manager. Thus, the accurate and efficient online schedulability test is essential for any task set whose periods are chosen from the QoS period set. For this purpose, we propose new utilization bounds as a function of given QoS periods. As long as there is a set of given periods that can be chosen by applications, the bounds developed can be used by the QoS manager to quickly determine the schedulability of dynamic applications.

Cache memory is used in almost all computer systems today to bridge the ever increasing speed gap between the processor and main memory. However, its use in multitasking computer systems introduces additional preemption delay due to the reloading of memory blocks that are replaced during preemption. This cache-related preemption delay poses a serious problem in realtime computing systems where predictability is of utmost importance. We propose an enhanced technique for analyzing and thus bounding the cache-related preemption delay in fixed-priority preemptive scheduling focusing on instruction caching. The proposed technique improves upon previous techniques in two important ways. First, the technique takes into account the relationship between a preempted task and the set of tasks that execute during the preemption when calculating the cache-related preemption delay. Second, the technique considers the phasing of tasks to eliminate many infeasible task interactions. These two features are expressed as constraints of a linear programming problem whose solution gives a guaranteed upper bound on the cache-related preemption delay. This paper also compares the proposed technique with previous techniques using randomly generated task sets. The results show that the improvement on the worst-case response time prediction by the proposed technique over previous techniques ranges between 5 percent and 18 percent depending on the cache refill time when the task set utilization is 0.6. The results also show that as the cache refill time increases, the improvement increases, which indicates that accurate prediction of cache-related preemption delay by the proposed technique becomes increasingly important if the current trend of widening speed gap between the processor and main memory continues.

In multi-tasking real-time systems, inter-task cache interference due to preemptions degrades schedulability as well as performance. To address this problem, we propose a novel scheduling scheme, called limited preemptive scheduling (LPS), that limits preemptions to execution points with small cache-related preemption costs. Limiting preemptions decreases the cache-related preemption costs of tasks but increases blocking delay of higher priority tasks. The proposed scheme makes an optimal trade-off between these two factors to maximize the schedulability of a given task set while minimizing cache-related preemption delay of tasks. Experimental results show that the LPS scheme improves the maximum schedulable utilization by up to 40\% compared with the traditional fully preemptive scheduling (FPS) scheme. The results also show that up to 20\% of processor time is saved by the LPS scheme due to reduction in the cache-related preemption costs. Finally, the results show that both the improvement of schedulability and the saving of processor time by the LPS scheme increase as the speed gap between the processor and main memory widens.

Fair queueing algorithms based on Generalized Processor Sharing (GPS) provide sessions with instantaneous fair sharing of server capacity, as well as guaranteed service rate and delay. The instantaneous fair sharing distributes server capacity to currently back-logged sessions in proportion to their weights regardless of the amount of service that the sessions received in the past. From a long-term perspective, however, this type of sharing leads to a different quality of service to sessions even if they have the same weight, since it provides only the instantaneous capacity sharing without considering the accumulated service delay and bandwidth. To minimize such long-term unfairness, we propose a delay and bandwidth normalized scheduling model in which the concept of Value of Service (VoS) is defined from the aspects of both delay and bandwidth. Performance comparisons between the proposed algorithm and traditional fair queueing algorithms show that the proposed algorithm provides better long-term fairness among sessions and is more adaptive to sessions with dynamic traffic patterns without compromising the guarantees of service rate and delay.

In this paper, we propose an optimization approach for robust control design with end-to-end response time constraints in a multi-resource cyber-physical systems (CPS). We introduce a rigorous performance metric for robust system design from the control theoretic viewpoint. Then, we investigate the impact of end-to-end response time analysis techniques on the control performance. We show that the traditional per-job response time analysis significantly degrades the control performance when real-time tasks visit a resource multiple times. We demonstrate that we can meaningfully improve the control performance by adopting the recently-developed per-resource response time analysis. Our simulation results verify the effectiveness of the proposed co-design framework.

IEEE 802.15.4 is a global standard designed for emerging applications in low-rate wireless personal area networks (LR-WPANs). The standard provides nice features such as a beacon-enabled mode and guaranteed time slots for real-time data delivery. However, how to optimally operate those features is still an open issue. For the optimal operation of the features, this paper proposes a holistic optimization method that jointly optimizes three cross-related problems: (1) cluster-tree construction, (2) nodes' power configuration, and (3) duty-cycle scheduling. Our holistic optimization method finds the solution for the three problems such that all the real time packets can be delivered within their deadlines in the most energy-efficient way. Our simulation study shows that, comparing with existing methods, our holistic optimization can guarantee the on-time delivery of all real-time packets while significantly saving the energy and hence significantly increasing the network lifetime.

Scheduling periodic tasks that allow some instances to be skipped produces spare capacity in the schedule. Only a fraction of this spare capacity is uniformly distributed and can easily be reclaimed for servicing aperiodic requests. The remaining fraction of the spare capacity is non-uniformly distributed, and no existing technique has been able to reclaim it. We present a method for improving the response times of aperiodic tasks by identifying the non-uniform holes in the schedule and adding these holes as extra capacity to the capacity queue of the CASH mechanism. The non-uniform holes can account for a significant portion of spare capacity, and reclaiming this capacity results in considerable improvements to aperiodic response times.

Monitoring network status such as end-to-end delay, jitter, and available bandwidth is important to support QoS-sensitive applications and timely detection of network anomalies like denial of service attacks. For this purpose, Internet service providers (ISPs) have started to instrument their networks with network measurement infrastructures (NMIs) that periodically run active measurement tasks using measurement servers located at strategic points in their networks. However, one problem that most network engineers have overlooked is the measurement conflict problem. Since active measurement tasks actively inject test packets to collect measurements along network paths, running multiple active measurements at the same time over the same path could result in misleading reports of network performance. We call this phenomenon a measurement conflict. Our recent observation of such measurement conflict motivates us to form a measurement task scheduling problem of meeting periodicity requirements, where real-time scheduling algorithms can play a role. The scheduling problem, however, is not exactly same as any of the existing scheduling problems in the realtime literature, because the problem involves multiple measurement servers running multiple measurement tasks whose conflict dependency propagates along the chains of paths. For this problem, we propose to use an EDF (earliest deadline first) heuristic but allowing "concurrent executions" if possible, to construct an offline schedule for a given measurement task set. Also, we propose a novel mechanism to flexibly use the offline schedule for minimizing the response time of dynamic on-demand measurement jobs. Further, we implement and deploy our scheduling algorithms in a real working NMI for monitoring Internet 2 Abilene network

This paper proposes a novel task model in which its physical and temporal parameters are specified as time-parameterized functions and their values are finally determined at the actual dispatch time. This model is clearly differentiated from the classical task model where parameters are fixed at the job release time. The new model better suits sensing tasks in tracking applications, since the sensor parameters such as field-of-view and measurement duration can be properly adjusted at the actual sensing time. The new model, however, creates the cyclic dependency between task parameters and scheduling behavior, that is, the task parameters depend on scheduling behavior and the latter in turn depends on the former. This cyclic dependency makes the schedulability check even more difficult. We handle this difficulty by iterative convergence and probabilistic schedulability envelope, which provides an efficient online schedulability check. The experimental study shows that the new model significantly improves the effective capacity of tracking systems without losing track accuracy

In order to satisfy and maintain service level agreements (SLAs), which demand high network availability and good network health, ISPs have started instrumenting their networks with network measurement infrastructures (NMIs) that are composed of dedicated measurement servers. Active measurements are frequently used in NMIs to regularly monitor network health and analyze the experience of end-user application traffic traversing the network. However, active measurements initiated by measurement servers need to be regulated. Unregulated active measurement traffic can cause an unpredictable negative impact on the actual application traffic. Also, running simultaneous conflicting active measurements on measurement servers could result in misleading reports of network performance. In this paper, we describe our active measurements scheduling framework called "OnTimeMeasure" that allows ISPs to regulate the amount of active measurement traffic injected into the network and also prevents conflicts in ongoing active measurements between measurement servers. OnTimeMeasure provides a simple scripting language interface to specify various measurement requirements such as physical topology of measurement server clusters, periodicity of the measurements, and properties of measurement tools. For a given measurement requirements script, OnTimeMeasure uses an efficient heuristic bin-packing algorithm to generate measurement timetables for orchestrating active measurements for a network involving multiple measurement servers, each hosting multiple measurement tools.

As the elderly population increases, the elderly care using inexpensive technological means becomes critical. This paper proposes novel scheduling algorithms for real-time indoor tracking of elderly residents, which is essential to assist and secure their independent living. Our scheduling algorithms are designed by harmonizing both sensing and communication signals and leveraging location-awareness and mobility-consciousness, in order to improve the tracking accuracy while reducing the energy consumption. We performed extensive experiments through both simulation and actual implementation. Our experimental result says that our scheduling algorithms can provide real-time tracking of residents within 20 cm error bound in the typical range of human mobility.

In recent studies, using mobile elements (MEs) as mechanical carriers of data has been shown to be an effective way of prolonging sensor network life time and relaying information in partitioned networks. As the data generation rates of sensors may vary, some sensors need to be visited more frequently than others. In this paper, a partitioning-based algorithm is presented that schedules the movements of MEs in a sensor network such that there is no data loss due to buffer overflow. Simulation results show that the proposed Partitioning Based Scheduling (PBS) algorithm performs well in terms of reducing the minimum required ME speed to prevent data loss, providing high predictability in inter-visit durations, and minimizing the data loss rate for the cases when the ME is constrained to move slower than the minimum required ME speed.

Resource reservation and QoS negotiation is a common way to guarantee timely progress of programs in distributed multimedia systems. For this, determining the available resource capacity, resource budget, is important. The resource budget depends on resource characteristics (e.g., processor, memory, disk, and network bandwidth) and scheduling algorithms. The paper provides an improved processor budget for the fixed-priority scheduling algorithm, which is most common in commercial real-time operating systems. The improvement is possible by noting that, in multimedia systems, there is a prefixed set of task periods for the finite set of QoS options and parameters. Our approach explicitly takes these periods into account and calculates the tight bound of the processor budget using the linear programming technique. This bound significantly improves Liu and Layland bound (Liu and Layland, 1973) and also it is proved to be better than any other bounds in the literature. We also show how this bound is effectively used for resource reservation and QoS re-negotiation for adapting to dynamic workload.

We propose an enhanced technique for analyzing, and thus bounding cache related preemption delay in fixed priority preemptive scheduling focusing on instruction caching. The proposed technique improves upon previous techniques in two important ways. First, the technique takes into account the relationship between a preempted task and the set of tasks that execute during the preemption when calculating the cache related preemption delay. Second, the technique considers phasing of tasks to eliminate many infeasible task interactions. These two features are expressed as constraints of a linear programming problem whose solution gives a guaranteed upper bound on the cache related preemption delay. The paper also compares the proposed technique with previous techniques. The results show that the proposed technique gives up to 60% tighter prediction of the worst case response time than the previous techniques.

In multi-tasking real-time systems, inter-task cache interference due to preemptions degrades system performance and predictability, complicating system design and analysis. To address this problem, we propose a novel scheduling scheme, called LPS (Limited Preemptible Scheduling), that limits preemptions to predetermined points with small cache-related preemption costs. We also give an accompanying analysis method that determines the schedulability of a given task set under LPS. By limiting preemption points, the proposed LPS scheme reduces preemption costs and thus increases the system throughput. Experimental results show that LPS can increase schedulable utilization by more than 10 % and save processor time by up to 44 % as compared with a traditional fully preemptible scheduling scheme.

Generalised Rate Monotonic Scheduling (GRMS) theory has now been widely adopted in practice and supported by open standards. In recent years, Ada runtime and COTS RTOSs supporting GRMS have met the DO 178B flight control standard (http://www.ddci.com/ products_darts.shtml; http://ic.arc.nasa.gov/publications/pdf/ 2000-0213.pdf). This creates strong incentives for the avionics industry to migrate from a traditional cyclical executive based system called a federated architecture to a GRMS based system to take advantage of GRMS's flexibility, and theoretical schedulability analysis. However, the industry is reluctant to migrate due to enormous potential re-certification cost. This paper presents a novel software architecture called a Logical Federated Architecture (LFA) that greatly reduces re-certification cost due to the change of scheduling methods.

This paper describes a stochastic analysis method for general periodic real-time systems. The proposed method accurately computes the response time distribution of each task in the system, thus making it possible to determine the deadline miss probability of individual tasks, even for systems with maximum utilization factor greater than one. The method uniformly covers both fixed-priority scheduling (such as rate monotonic) as well as dynamic-priority scheduling (such as earliest deadline first) and can handle arbitrary relative deadlines and execution time distributions. The accuracy of the method is proven by comparing the results from the analysis with those obtained from simulations, as well as other methodologies in the literature.