Online dispatching and parallel processing algorithms for saving money in systems with heterogeneous, single-buffered, speed-scalable processors

We investigated attention allocation in a dual-task paradigm using behavioral and pupillary measures. We used an auditory digit span (DS) and a simple visual response time (RT) task. Participants were administered four conditions in which they performed neither task (no-task), a single task (DS or RT only), or both tasks (dual). Dependent variables were DS accuracy, RT, and task-evoked pupillary responses (TEPRs) to digits as estimates of mental effort. Participants maintained almost the same level of DS accuracy on dual as on DS only and sacrificed speed on the RT task. As expected, TEPRs increased linearly with memory load in both DS only and dual. Although TEPRs were initially higher in dual than in DS only, the slope of the increase was shallower in dual. Results suggest that TEPRs can elucidate mechanisms of attention allocation by distinguishing between effectiveness (level of behavioral performance) and efficiency (the costs of that performance in mental effort).

In scheduling algorithms based on the Rate Monotonic (RM) method which are widely used in development of real-time systems, tasks with shorter periods have higher priorities. In contrast, ones with longer periods are likely to suffer from increased response times and jitter due to their lower priorities. We propose an Execution Right Delegation (ERD) method based on RM where a high-priority server for particular (or important) task is introduced to shorten response time and jitter of the task. In the evaluation, it is confirmed that response times and jitter of a particular (important) task are reduced. We also show Response Time Analysis (RTA), which assures worst-case response time of the task. This paper shows the algorithm and RTA of ERD and evaluates it by comparing it to a Deadline Monotonic method. The evaluation by simulation shows that ERD can reduce the average worst-case response time by 13.45% at maximum compared to the Deadline Monotonic scheduling. In addition, we confirm that the RTA provides a worst-case response time close to the simulation results.

Response time is affected by interprocessor communications, precedence relationships among the modules, module assignments, and processor scheduling policies. Furthermore, due to sharing of resources and data among the processors, contention delays are incurred. A task response time model that considers all these factors is proposed. A Petri net is used to represent resource contention, and the task control flow graph represents module precedence and logical relationships. A queuing network with resource contention is used to estimate the response time of each module. Module response time consists of delays at the processors and resource queues and is estimated by approximating the extended queuing network as independent finite capacity queuing systems. The module response time is mapped onto a control flow graph, and task response time is obtained by aggregating the module response times in accordance with their precedence relationship in the control flow graph. The task response time derived from the analytical model compares well with that from the simulation. >

In a real-time system with both hard deadline periodic tasks and soft deadline aperiodic tasks, hard deadline tasks must be guaranteed to meet their deadlines and soft deadline tasks are serviced in a best-effort manner. To schedule soft deadline aperiodic tasks in dynamic priority systems, one can assign a virtual deadline like the TBS algorithm or use capacity servers like the DPE algorithm. Both approaches have their own advantages, and we cannot say which one is better in terms of response time. In some cases, TBS provides shorter response time than DPE server. In other cases, the reverse may happen. In this paper, we propose a combined approach to improve the response time of soft deadline aperiodic tasks. The proposed method combines the deadline assignment and the capacity server approach. When there is enough capacity available for aperiodic tasks, the aperiodic tasks are served by capacity servers. Otherwise, our method assigns a deadline and schedule aperiodic task together with periodic tasks by EDF algorithm. In the worst case, the proposed method provides response time no later than TBS. Experimental results show that the proposed approach improves the responsiveness of soft deadline aperiodic tasks.

Mobile Edge Computing is a new technology which aims to reduce latency, to ensure highly efficient network operation and to offer an improved user experience. Considering offloading will introduce additional wireless transmission overhead, the key technical challenge of mobile edge computing is tradeoff between computation cost and wireless transmission cost, reducing energy consumption of mobile edge devices and response time of computation task at the same time. A Mobile Edge Computing System composed of Mobile Edge Device and Edge Cloud, connecting with Wireless Stations, comes out. To protect user privacy, Data Preprocessing is proposed which includes irrelevant property clean and data segmentation. Aimed at reducing total energy consumption and response time, an energy consumption priority offloading (ECPO) algorithm and a response time priority offloading (RTPO) algorithm are put forward, based on Energy Consumption Model and Response Time Model. Combining both ECPO and RTPO, a dynamic computing offloading algorithm is raised which is more universal. Finally, simulations in four scenarios, including network normal scenario, network congested scenario, device low battery scenario and task time limited scenario, demonstrate that our algorithms can effectively reduce energy consumption of mobile edge device and response time of computation task.

Many modern parallel applications, such as desktop software and cloud-based web services, are service-oriented, long running, and perform frequent interactions with the external world (e.g., responding to user input). We want such interactive applications to provide fast response times because typically at the other end of the external interaction there is a user waiting for a response. Existing parallel platforms designed for multicore hardware do not work well for such interactive applications, because they are designed to maximize throughput (rather than responsiveness). Interactive applications may have a mixture of interactive and compute-intensive tasks occurring concurrently, and the scheduler must be able to discern and prioritize tasks so that tasks which require faster response are prioritized over background tasks. We present Interactive Cilk, or I-Cilk for short, a task parallel platform designed to schedule such parallel interactive applications. I-Cilk supports a C++-based templated library that allows the programmer to specify priorities for task-parallel code, and the underlying runtime schedules the computation so as to optimize for the response time of high-priority tasks. We show that the scheduling algorithm used by I-Cilk provides provably efficient response times for tasks at all levels of priorities, with better response time to high-priority tasks. We also empirically demonstrate that the scheduling algorithm can be implemented efficiently in practice with low scheduling overhead and provides fast response times for high-priority tasks.

Real-time applications may be comprised of both soft and hard tasks, the former being related to streaming-based or other non-critical services. For these applications one is interested in minimizing the response time of soft tasks, for which quality of service is at stake, while not jeopardizing hard deadlines. In this paper, we propose a framework for multiprocessor real-time systems, that can schedule hard and soft tasks with unconstrained deadlines according to two EDF-based configurations. Our solution relies on two special servers, which are responsible for providing temporal isolation and slack reclaiming. These strategies are combined so that the average response time of soft tasks are improved. Results are evaluated via extensive simulation, which indicates that: the available processing capacity can be effectively utilized; average response time of aperiodic tasks is significantly reduced and hard deadlines are preserved while soft deadline miss ratio is kept low.

Previous studies demonstrate that semantics, the higher level meaning of multi-modal stimuli, can impact multisensory integration. Valence, an affective response to images, has not yet been tested in non-priming response time (RT) or temporal order judgement (TOJ) tasks. This study aims to investigate both semantic congruency and valence of non-speech audiovisual stimuli on multisensory integration via RT and TOJ tasks (assessing processing speed (RT), point of subjective simultaneity (PSS), and time-window when multisensory stimuli are likely to be perceived as simultaneous (Temporal Binding Window; TBW)). Forty participants (mean age: 26.25; females=17) were recruited from Prolific Academic resulting in 37 complete datasets. Both congruence and valence have a significant main effect on RT (congruent and high valence decrease RT) as well as an interaction effect (congruent/high valence condition being significantly faster than all others). For TOJ, images high in valence require visual stimuli to be presented significantly earlier than auditory stimuli in order for the audio and visual stimuli to be perceived as simultaneous. Further, a significant interaction effect of congruence and valence on the PSS revealed that the congruent/high valence condition was significantly earlier than all other conditions. A subsequent analysis shows there is a positive correlation between the TBW width (b-values) and RT (as the TBW widens, the RT increases) for the categories that differed most from 0 in their PSS (Congruent/High and Incongruent/Low). This study provides new evidence that supports previous research on semantic congruency and presents a novel incorporation of valence into behavioural responses.

Most of the braille embossers incorporate a mode for the creation of braille embossed graphics (BEG). Some of them can produce dots in different elevations, but also various dot densities. Dot elevation and dot density constitute the main variables that dominate the identification and conceptualization by blind users of tactile elements, such as lines and areas, in BEG. Nevertheless, there are not enough published research data in this domain. We present the results of a study with 30 congenitally blind participants to examine their detection performance in tactile identification of context-free embossed lines and square areas in eight dot elevations H1 to H8 (from 0.03 to 0.47 mm) and two dot densities (10 and 20 dpi). The detection accuracy and response time of the matching task were examined for each type of test stimuli. The results of correct and misclassified stimuli are presented in confusion matrixes, along with the overall mean response time for each task. The participants classify with better accuracy the stimuli with the three lower dot elevations (between 0.03 mm and 0.15 mm) for both types of stimuli and in both dot densities. The minimum overall mean response time of the identification task was achieved for the larger dot height H8 in the case of raised-dot square areas of 10 dpi. The analysis of the results indicates that better classification accuracy is achieved with the combination of four dot heights H1, H2, H4, and H8. Blind users can distinguish textures produced by dots with very low heights in line with the recent literature suggestions. The current study constitutes a first step toward recommendations for the designers of BEG in applying textures with variable dot height and dot density for educational purposes or for producing tactile maps to benefit users who are blind.


 Objectives: The main objective of this study was to evaluate the laterality recognition of the body parts “hands”, “neck” and “feet” based on the response time and accuracy of left/right body-part judgments with performance compared across three age groups. The secondary objective was to calculate within age correlations between the time and accuracy variables and a measure of physical activity that provides activity in metabolic equivalents [METs]), and the METs score.
 Methods: A total sample of 154 healthy subjects was divided into three age groups: G1 (n = 58), G2 (n = 38), and G3 (n = 58). Response time and accuracy were measured with the application Recognise designed by the NOI Group for “hand”, “neck” and “feet” images.
 Results: The results showed significant changes between groups in accuracy and response time for all body parts. Post-hoc analysis revealed longest response time and worst accuracy for G3. In the correlation with METs analysis, the strongest correlations were found in G1, where there was a positive moderate association between the METs score and the accuracy for the right-hand images (r = 0.352) and right neck images (r = 0.398).
 Conclusion: Body-part laterality recognition is altered with age. Longer response time and lower accuracy was observed for “hands”, “neck” and “feet”.

In hard real-time task systems where periodic and aperiodic tasks coexist, the object of task scheduling is to reduce the response time of the aperiodic tasks while meeting the deadline of periodic tasks. Total bandwidth server (TBS) and advanced TBS (ATBS) are used in dynamic priority systems. However, these methods are not optimal solutions because they use the worst-case execution time (WCET) or the estimation value of the actual execution time of the aperiodic tasks. This paper presents an online slack-stealing algorithm called SSML that can make significant response time reducing by modification of look-ahead earliest deadline first (laEDF) algorithm as the slack computation method. While the conventional slack-stealing method has a disadvantage that the slack amount of each frame must be calculated in advance, SSML calculates the slack when aperiodic tasks arrive. Our simulation results show that SSML outperforms the existing TBS based algorithms when the periodic task utilization is higher than 60%. Compared to ATBS with virtual release advancing (VRA), the proposed algorithm can reduce the response time up to about 75%. The performance advantage becomes much larger as the utilization increases. Moreover, it shows a small performance variation of response time for various task environments.

We propose power-saving real-time scheduling method for mixed task sets which consist of both time-based periodic and event-based aperiodic tasks in the automotive operating system. In order to consider power saving, it is necessary to estimate the slack time and minimize the response time of the aperiodic task simultaneously. However, because these two goals conflict with each other, It must compromise between the two goals, depending on the given application domain. In this paper, we find an adjustment factor that improves the response time of the aperiodic task with a slight increase in power consumption. We propose the lccEDF algorithm and balance the two goals by systematically adjusting the factors. Our simulation results show that our approach is excellent in various task sets.

In a series of experiments, we examined the effect of requiring subjects to attend to distractors in a test of negative priming. This was accomplished by using a referent size-selection task in which subjects were instructed to name the larger animal and to ignore the smaller animal in a word pair. The result was a quadrupling of the standard negative priming effect, suggesting that negative priming not only occurs for attended distractors, it is actually enhanced. We demonstrated that this enhancement of the effect was not due solely to increased latencies in the referent size-selection task, because neither decreasing base response times in other referent size-selection tasks nor increasing base response times in typical color-selection tasks substantially affected the respective negative priming effects. Although these findings can be accommodated within current theories of negative priming, they challenge the basic assumption that the negative priming effect arises because the critical item was ignored or not attended to on the prime trial.

Real-time embedded systems are characterized by their need to complete task executions within their scheduled deadlines. Task scheduling could get complicated due to the occurrence of faults in such systems. Traditionally, researchers have accounted for faults by assuming a constant number of faults occuring in the system. However, such an assumption may lead the system designers to rather pessimistic estimates on tasks' response times. In this paper, the occurrence of faults has been modeled as a stochastic process with a Poisson distribution having a mean inter-arrival rate of lambda. The usefulness of the model has been evaluated in terms of estimated task response time, energy consumption and loss of predictability for two periodic real-time embedded task sets, one real-world and another synthetic. It has been shown that the proposed approach can achieve up to 40% improvement in terms of the estimated energy savings and task response time with very minimal loss of predictability in the system of about 1e-3 % as compared to the systems with constant number faults

In many real-world systems incoming tasks split into subtasks which are processed by a set of parallel servers. In such systems two metrics are of potential interest: response time and subtask dispersion. Previous research has been focused on the minimisation of one, but not both, of these metrics. In particular, in our previous work, we showed how the processing of selected subtasks can be delayed in order to minimise expected subtask dispersion and percentiles of subtask dispersion in the context of split-merge systems. However, the introduction of subtask delays obviously impacts adversely on task response time and maximum sustainable system throughput. In the present work, we describe a methodology for managing the trade off between subtask dispersion and task response time. The objective function of the minimisation is based on the product of expected subtask dispersion and expected task response time. Compared with our previous methodology, we show how our new technique can achieve comparable subtask dispersion with substantial improvements in expected task response time.