Variable Execution Times Research Articles

A Novel On-Chip Task Scheduler for Mixed-Criticality Real-Time Systems

This paper presents a novel design of a coprocessor that performs hardware-accelerated task scheduling for embedded real-time systems consisting of mixed-criticality real-time tasks. The proposed solution is based on the Robust Earliest Deadline (RED) algorithm and previously developed hardware architectures used for scheduling of real-time tasks. Thanks to the HW implementation of the scheduler in the form of a coprocessor, the scheduler operations (i.e., instructions) are always completed in two clock cycles regardless of the actual or even maximum task amount within the system. The proposed scheduler was verified using simplified version of UVM and applying billions of randomly generated instructions as inputs to the scheduler. Chip area costs are evaluated by synthesis for Intel FPGA Cyclone V and for 28-nm TSMC ASIC. Three versions of real-time task schedulers were compared: EDF-based scheduler designed for hard real-time tasks only, GED-based scheduler and the proposed RED-based scheduler, which is suitable for tasks of various criticalities. According to the synthesis results, the RED-based scheduler consumes LUTs and occupies larger chip area than the original EDF-based scheduler with equivalent parameters used. However, the RED-based scheduler handles variations of task execution times better, achieves higher CPU utilization and can be used for the scheduling of hard real-time, soft real-time and nonreal-time tasks combined in one system, which is not possible with the former algorithms.

Towards Heterogeneous Multi-scale Computing on Large Scale Parallel Supercomputers

New applications that can exploit emerging exascale computing resources efficiently, while providing meaningful scientific results, are eagerly anticipated. Multi-scale models, especially multi-scale applications, will assuredly run at the exascale. We have established that a class of multi-scale applications implementing the heterogeneous multi-scale model follows, a heterogeneous multi-scale computing (HMC) pattern, which typically features a macroscopic model synchronising numerous independent microscopic model simulations. Consequently, communication between microscopic simulations is limited. Furthermore, a surrogate model can often be introduced between macro-scale and micro-scale models to interpolate required data from previously computed micro-scale simulations, thereby substantially reducing the number of micro-scale simulations. Nonetheless, HMC applications, though versatile, remain constrained by load balancing issues. We discuss two main issues: the a priori unknown and variable execution time of microscopic simulations, and the dynamic number of micro-scale simulations required. We tackle execution time variability using a pilot job mechanism to handle internal queuing and multiple sub-model execution on large-scale supercomputers, together with a data-informed execution time prediction model. To dynamically select the number of micro-scale simulations, the HMC pattern automatically detects and identifies three surrogate model phases that help control the available and used core amount. After relevant phase detection and micro-scale simulation scheduling, any idle cores can be used for surrogate model update or for processor release back to the system. We demonstrate HMC performance by testing it on two representative multi-scale applications. We conclude that, considering the subtle interplay between the macroscale model, surrogate models and micro-scale simulations, HMC provides a promising path towards exascale for many multiscale applications.

Formulation of Control Strategies for IoT Task Scheduling

The various Internets of Things (IoT) application tasks are difficult to schedule due to heterogeneity properties of IoT. So an efficient algorithm is required that forms < task, processor> pair appropriately. This paper presents a more sensible model for varying execution times of tasks and deviation in task parameters for building a schedule is allowed. The system provides an adaptive learning mechanism called Expected Time Matrix ETM (i, j). When the environment of the system changes dynamically, the system learns and adapts itself to the new changes automatically, since the learning mechanism has been incorporated in the system. ETM (i, j) concepts allows the system to learn from past instances as well. The work is supported by simulations that highlight the viability of concepts proposed. The key objective of this paper is to present the developed scheduling algorithm that is self-configurable and dynamic

Power of lower extremities and age were the main determinants on the agility test for adults in a cohort of men aged 66–91 years

Study Design Cross-sectional study. Objective To evaluate the relationship between agility and personal factors, muscle strength and power, mobility, self-reported balance and physical activity among older men. Methods Agility was measured by using the Agility Test for Adults (ATA). We studied 100 Finnish male former elite athletes (endurance n = 50; power n = 50) and 50 matched controls aged 66 to 91 years (mean age 75.5 years). The associations between agility and other variables were similar between three groups; thus, multiple linear regression analyses were done by using the pooled data of the participants. Results On the basis of multiple linear regression analyses, combination of age (p = .02), self-reported Activities-specific Balance Confidence scale (ABC scale), jumping height (p = .001) and self-rated health explained 26% of the variance in execution time of ATA (R2 = 0.26; p = .000002) among elderly men. Conclusion Power of lower extremities and age were the main determinants of the results of ATA in a cohort of men aged 66–91 years. From a clinical point of view, power of lower extremities measured by test demanding explosive power plays an important role to maintain or enhance capacity of agility.

Geometric Accuracy Study of Orthorectification Based on Sensor Model Refinement in Imagery Subset Using ORFEO Toolbox (OTB)

Large scale base map can be obtained by various methods, one of them is orthorectification process of remote sensing satellite imagery to eliminate the relief displacement caused by height variation of earth surface. To obtain a map images with good quality, it requires additional data such as sensor model in the form of rational polynomial coefficients (RPC), surface model data, and ground control points Satellite imageries with high resolution file size are relatively large. In order to process them, high specification of hardwares were required. To overcome this by cutting only a portion of the images, based on certain study areas were suffer from of georeference lost so it would not be able to orthorectified. On the other hand, in several remote sensing software such as ESA SNAP and Orfeo Toolbox (OTB) subset or pixel extraction from satellite imagery, preserve the imagery geometric sensor models. This research aimed at geometric accuracy of orthorectification carried out in a single scene of Pleiades Imagery within the Kepahiang Subdistrict, located at Kepahiang Regency, Bengkulu Province, by using DEMNAS and the imagery refined sensor mode, and ground control points taken using GPS Survey. Related with the raw imagery condition which consists of Panchromatic and multispectral bands, this study were separated to assembly, pan sharpening , and sensor model refinement stages prior to orthorectification carried out both in the original or full extent imagery and the result of subset extent imagery. After these processses taken place, we measure the accuracy of each full and subset imagery.These procedures were carried out using Orfeo toolbox 6.6.0 in the Linux Mint 19 Operating system. From the process log, running time in total were 7814.518 second for the full extent and 4321.95 seconds for the subset processess. And as a big data process, the total of full extent imageries was 83.15 GB while the subset size was only 30.73 GB. The relative accuracy of the full extent and its subset imagery were 0.431 meters. Accuracy of the sensor model refinement process are 1.217 meters and 1.550 meters with GCP added, while the accuracu of the orthorectifications results were 0.416 meters and 0.751 meters by using ICP. Variation of execution time may caused by the data input size and complexity of the mathematical process carried out in each stages. Meanwhile, the variation of accuracy may caused by the check or control points placements above satellite Imagery which suffer from uncertainty when dealing with the sub-pixel position or under 0.5 meters.

Age-related kinematic performance should be considered during fast head-neck rotation target task in individuals aged from 8 to 85 years old.

Kinematic behavior during fast cervical rotations is a useful parameter for assessing sensorimotor control performances in neck-pain patients. However, the influence of age in asymptomatic individuals from children to older people still needs to be explored. Our aim was to assess the impact of age on sensorimotor control performance of the head-neck with execution time and kinematic variables (time of task, mean speed/acceleration/deceleration, overshoots (OSs), minimum/maximum speed) during standardized fast rotation target task using the DidRen Laser test. A total of 80 volunteers were stratified in four different age-groups: Children (8–14 years): n = 16; Young Adults (18–35 years): n = 29; Old Adults (36–64 years): n = 18; Seniors (65–85 years): n = 17. Results showed that to perform the test, Children were slower (69.0 (60.6–87.3)s) compared to Young Adults (49.6 (45.6–55.6)s) with p < 0.001, and Old Adults (51.7 (48.4–55.8)s) with p < 0.001. It was also slower in Seniors (57 (52.3–67.6)s) compared to Young Adults with p < 0.013. Mean speed was slower in Children (9.4 ± 2.3 °s−1) and Seniors (10.6 ± 2.4 °s−1) compared to Young Adults (13.7 ± 1.9 °s−1) with p < 0.001 and Old Adults (13.3 ± 2.4 °s−1) with p < 0.001. Mean acceleration was slower for Children (8.4(7.6–10.2) °s−2) compared to Young Adults (11.1 (8.8–15.3) °s−2) with p < 0.016, and Old Adults (12.0(8.4–15.3) °s−2) with p < 0.015. Mean deceleration was slower for Children (−1.9(−2.6–1.4) °s−2) compared to Young Adults (−2.9(−3.7–2.5) °s−2) with p < 0.001 and Old Adults (−3.2(−3.7–2.3) °s−2) with p < 0.003. The DidRen Laser test allows us to discriminate age-specific performances for mean speed, acceleration and deceleration. Seniors and Children needed to be slower to become as precise as Young Adults and Old Adults. No difference was observed for OSs which assesses accuracy of movement. Age should therefore be considered as a key parameter when analyzing execution time and kinematic results during DidRen Laser test. These normative data can therefore guide clinicians in the assessment of subjects with neck pain.

NEWS: Optimizing the Service Performance for Workflows with Replications in Virtualized Cloud Environments

Cloud computing environments facilitate applications by providing virtualized resources. It is attracting increasing interests for its elasticity and high quality of service, especially in “big time”. For applications, execution time is crucial for the quality of service, because works finished after the deadlines will be useless. Nevertheless, typical cloud environments do not present regular performance in terms of execution and data transmission. And cloud service providers cannot guarantee the performance of execution time, which would take large default risk. In this paper, a performaNce guaranteE Workflow Scheduling (NEWS) algorithm is proposed to address these issues by taking replications based on reconstructed workflow. Characteristics of these issues are summarized to further describe the problems and challenges, which have always been ignored. The algorithm employs the parameter of partial critical path to optimize the workflow scheduling performance. On the analysis of challenges brought up by taking replications, NEWS also optimizes the choice of tasks and the number of replications. Simulation experiments show that the proposed algorithm reduces the variation of total execution time, which makes it possible to provide performance guarantee services.

A New Standardization of the Bells Test: An Italian Multi-Center Normative Study.

Objective: The Bells Test is a cancelation task that is widely used for the diagnosis of unilateral spatial neglect (USN). With the aim of fostering more reliable use of this instrument, we set out to develop new norms adjusted for the possible influence of age, gender and education. We worked on the original version of the test.Methods: Normative data were collected from 401 healthy participants aged between 20 and 80 years. Individual factors that could affect performance (i.e., gender, age, and years of education) were considered. We computed several indices on the Bells Test including an asymmetry score, an accuracy score and execution time. Multiple regression analyses (for time measures) and generalized linear models (for accuracy measures) were used to check for the influence of individual predictors of performance on the Bells Test.Results: Data indicated a significant influence of age on the accuracy score and execution time variables and a marginally significant effect of education on the accuracy score variable. Wherever appropriate, cut-offs are provided for the three dependent scores on the Bells Test corrected for age and education.Conclusion: Based on a large normative sample, the present study provides new normative data on the Bells Test, which could lead to its reliable use in the diagnosis of USN.

Using simulation to reshape the maintenance systems of caster segments

The maintenance of steel mills has a relevant impact on costs and productivity of steel industry. According to the traditional approach, one crew per shift takes care of the maintenance of key components of the slab caster with very limited process control, productivity and long lead-times. This paper introduces a new approach that makes each component move along a line where specific attention has to be paid to shared resources. The methodological approach is based on simulation, which has been applied to a real-life case study, represented by a leading Indian steel mill. Simulation is exploited, first, to point out the effect of the concurrent requirement of resources, demonstrating that a static sizing can lead to wrong designs, then, to properly consider the variance of execution times. Results show that caster maintenance arranged in a line designed according to the proposed methodology remarkably contributes to speed up the maintenance process.

Using simulation to reshape the maintenance systems of caster segments

The maintenance of steel mills has a relevant impact on costs and productivity of steel industry. According to the traditional approach, one crew per shift takes care of the maintenance of key components of the slab caster with very limited process control, productivity and long lead-times. This paper introduces a new approach that makes each component move along a line where specific attention has to be paid to shared resources. The methodological approach is based on simulation, which has been applied to a real-life case study, represented by a leading Indian steel mill. Simulation is exploited, first, to point out the effect of the concurrent requirement of resources, demonstrating that a static sizing can lead to wrong designs, then, to properly consider the variance of execution times. Results show that caster maintenance arranged in a line designed according to the proposed methodology remarkably contributes to speed up the maintenance process.

Efficient Fine-grained Processor-logic Interactions on the Cache-coherent Zynq Platform

The introduction of cache-coherent processor-logic interconnects in CPU-FPGA platforms promises low-latency communication between CPU and FPGA fabrics. This reduced latency improves the performance of heterogeneous systems implemented on such devices and gives rise to new software architectures that can better use the available hardware. Via an extended study accelerating the software task scheduler of a microkernel operating system, this article reports on the potential for accelerating applications that exhibit fine-grained interactions. In doing so, we evaluate the performance of direct and cache-coherent communication methods for applications that involve frequent, low-bandwidth transactions between CPU and programmable logic. In the specific case we studied, we found that replacing a highly optimised software implementation of the task scheduler with an FPGA-based scheduler reduces the cost of communication between two software threads by 5.5%. We also found that, while hardware acceleration reduces cache footprint, we still observe execution time variability because of other non-deterministic features of the CPU.

Slack clustering for scheduling frame-based tasks on multicore embedded systems

Adopting multicore platforms is a general trend in real-time embedded systems. However, integrating tasks with different real-time constraints into a single platform presents new design challenges. While it must be guaranteed that hard real-time tasks are able to meet their deadline even in worst case scenarios, firm real-time tasks should be scheduled in a way to achieve high system utilization in order to provide a better quality of service. In this paper, we propose a scheduling scheme for frame-based tasks on real-time multicore embedded systems which is able to guarantee the schedulability of the hard real-time tasks, while it improves the number of executed firm real-time tasks. Considering intrinsic variation of execution time of tasks and with the help of a clustering algorithm, hard real-time tasks are scheduled in a way that their expected slack time could be exploited more efficiently for the execution of firm tasks. The extensive simulations show that our proposed scheme can improve the system utility up to 60% compared to a state-of-the-art scheduling technique.

Optimal harmonic period assignment: complexity results and approximation algorithms

Harmonic periods have wide applicability in industrial real-time systems. Rate monotonic (RM) is able to schedule task sets with harmonic periods up to 100% utilization. Also, if there is no release jitter and execution time variation, RM and EDF generate the same schedule for each instance of a task. As a result, all instances of a task are interfered by the same amount of workload. This property decreases the jitters that happen during sampling and actuation of the tasks, and hence, it increases the quality of service in control systems. In this paper, we consider the problem of optimal period assignment where the periods are constrained to be harmonic and the task set is required to be feasible. We study two variants of this problem. In the first one, the objective is to maximize the system utilization, while in the second one, the goal is to minimize the total weighted sum of the periods. First, we assume that an interval is determined a priori for each task from which its period can be selected. We show that both variants of the problem are (at least) weakly NP-hard. This is shown by reducing the NP-complete number partitioning problem to the mentioned harmonic period assignment problems. Afterwards, we consider a variant of the second problem in which the periods are not restricted to a special interval. We present two approximation algorithms with polynomial-time complexity for this problem and show that the maximum relative error of these algorithms is bounded by a factor of 1.125. Our evaluations show that, on the average, results of the approximation algorithms are very close to an optimal solution.

Quantifying and reducing execution variance in STM via model driven commit optimization

Simplified parallel programming coupled with an ability to express speculative computation is realized with Software Transactional Memory (STM). Although STMs are gaining popularity because of significant improvements in parallel performance, they exhibit enormous variation in transaction execution with non-repeatable performance behavior which is unacceptable in many application domains, especially in which frame rates and responsiveness should be predictable. Thus, reducing execution variance in STM is an important performance goal that has been mostly overlooked. In this work, we minimize the variance in execution time of threads in STM by reducing non-determinism exhibited due to speculation by first quantifying non-determinism and generating an automaton that models the behavior of STM. We used the automaton to guide the STM to a less non-deterministic execution that reduced the variance in frame rate by a maximum of 65% on a version of real-world Quake3 game.

Differences in the Performance of Parallel Applications in Physical and Virtual Clusters

This article evaluates how software environment influences on the performance of parallel applications when they run in a dedicated, shared or virtual cluster, to identify the factors that influence performance according to the type of environment and the behavior of the applications. The evaluations were performed using the NAS benchmark software and the clusters were built on the same hardware configurations. The data obtained served to identify the factors involved in the variations of execution times between the different tested configurations. The results indicate that the performance of some applications is not always better in a physical cluster than in a virtual cluster, which a priori seems counterintuitive.

An Energy-Efficient Task Scheduling for Near-Realtime Systems with Execution Time Variation

An Energy-Efficient Task Scheduling for Near-Realtime Systems with Execution Time Variation

GPU-Accelerated Real-Time Path Planning and the Predictable Execution Model

Path planning is one of the key functional blocks for autonomous vehicles constantly updating their route in real-time. Heterogeneous many-cores are appealing candidates for its execution, but the high degree of resource sharing results in very unpredictable timing behavior. The predictable execution model (PREM) has the potential to enable the deployment of real-time applications on top of commercial off-the-shelf (COTS) heterogeneous systems by separating compute and memory operations, and scheduling the latter in an interference-free manner. This paper studies PREM applied to a state-of-the-art path planner running on a NVIDIA Tegra X1, providing insight on memory sharing and its impact on performance and predictability. The results show that PREM reduces the execution time variance to near-zero, providing a 3× decrease in the worst case execution time.

Balancing stochastic U-lines using particle swarm optimization

U-lines are important parts of the Just-In-Time production system in order to improve productivity and quality. In real life applications of assembly lines, the tasks may have varying execution times defined as a probability distribution. In this study, a novel particle swarm optimization algorithm is proposed to solve the U-line balancing problem with stochastic task times. A computational study is conducted to compare the performance of the proposed approach to the existing methods in the literature. The results of the computational study show that the proposed approach performs quite effectively. It also yields good solutions for all test problems within a short computational time.

A Survey of Techniques for Cache Locking

Cache memory, although important for boosting application performance, is also a source of execution time variability, and this makes its use difficult in systems requiring worst-case execution time (WCET) guarantees. Cache locking is a promising approach for simplifying WCET estimation and providing predictability, and hence, several commercial processors provide ability for locking cache. However, cache locking also has several disadvantages (e.g., extra misses for unlocked blocks, complex algorithms required for selection of locking contents) and hence, a careful management is required to realize the full potential of cache locking. In this article, we present a survey of techniques proposed for cache locking. We categorize the techniques into several groups to underscore their similarities and differences. We also discuss the opportunities and obstacles in using cache locking. We hope that this article will help researchers gain insight into cache locking schemes and will also stimulate further work in this area.

Multi-objective optimization of multimedia embedded systems using genetic algorithms and stochastic simulation

To meet the ever shrinking time-to-market for multimedia embedded systems, designers need effective system-level optimization techniques to support their design decisions. Despite multimedia embedded systems' highly variable execution times and soft real-time constraints, most previous work has adopted a constant execution time (worst-case) approach to evaluate if a candidate architecture satisfies the timing constraints. Such an approach is too pessimistic and might result in unnecessary costly architectures. In this work, we propose a new method for design space exploration of multimedia embedded systems. Given a system specification, the proposed method automatically explores the design space to quickly identify Pareto-optimal solutions (or an approximation) that optimize conflicting design metrics, such as price and power consumption. Our approach combines (i) a fast and formal strategy for performance evaluation that captures the varying runtime behavior of multimedia systems and (ii) a new multi-objective genetic algorithm for architecture exploration. The experiments on well-known benchmarks show the efficiency of our method in comparison to similar ones.