Response Time Estimation Research Articles

Basic Requirements to the SPIIRAS Transactions Paper Format Feasibility of Real-Time Applications on Multicore Processors

A brief survey of key results is presented on the feasibility analysis of real–time applications that are implemented by multicore processors with various scheduling methods. A method for estimation of task response time is suggested for scheduling with static task priorities. Number of planning points and number of context switches are estimated for quantum– based and non–quantum scheduling methods.

Estimation of blocking factor for tasks in real-time systems with multi-core processors

An approach is described for estimate of task response time in real-time systems that are implemented by multi-core processors. The fact is shown, that in systems on multi-core processors the composite blocking of tasks is possible not only with use of priority inheritance protocols, but also with use other known protocols of mutual data access. A method is given for estimate of blocking factor for systems with composite blocking of tasks on multi-core processors.

Response time estimation for task chains in destributed realtime systems

A method for estimation of time, requiered for external events processing by distributed program applications is presented. It is shown that feasibility of such applications may depend on variation of task execution time and message delivery time. The influence of this variation shall be taken into account during feasibility checking for task chain, that participate in reaction to concrete separate external events.

Evaluating Beat Structure and Truck Allocation for the Tarrant County, Texas, Courtesy Patrol

Nonrecurring congestion poses a significant concern for drivers on urban freeways. Effective incident management relies on many tools to prevent the delays caused by crashes, road debris, and stalled or disabled vehicles. Many urban areas have adopted freeway service patrol (FSP) programs that search the freeway network for incidents, provide aid to motorists, and assist with incident management and clearance. FSP management must consider the patrols' beat structure and the fleet allocation. Both deterministic and probabilistic response time estimates for each beat are used to assess fleet allocations. The goal is to consider whether the urban network should be segmented into as many beats as possible, with individual trucks assigned to each beat, or whether additional trucks should be allocated to fewer beats. The Tarrant County, Texas, Courtesy Patrol (CP) is used as a case study to explore the truck allocation problem with field data. The Tarrant County CP typically uses a one-beat, two-beat, or three-beat configuration with a single truck allocated to each beat. The merits of a two-beat configuration are examined. The study indicates that the incident distribution must be considered during both beat and tour design.

Using enterprise architecture analysis and interview data to estimate service response time

Insights into service response time is important for service-oriented architectures and service management. However, directly measuring the service response time is not always feasible or can be very costly. This paper extends an analytical modeling method which uses enterprise architecture modeling to support the analysis. The extensions consist of (i) a formalization using the Hybrid Probabilistic Relational Model formalism, (ii) an implementation in an analysis tool for enterprise architecture and (iii) a data collection approach using expert assessments collected via interviews and questionnaires. The accuracy and cost effectiveness of the method was tested empirically by comparing it with direct performance measurements of five services of a geographical information system at a Swedish utility company. The tests indicate that the proposed method can be a viable option for rapid service response time estimates when a moderate accuracy within 15% is sufficient.

Analytical Performance Models for MapReduce Workloads

MapReduce is a currently popular programming model to support parallel computations on large datasets. Among the several existing MapReduce implementations, Hadoop has attracted a lot of attention from both industry and research. In a Hadoop job, map and reduce tasks coordinate to produce a solution to the input problem, exhibiting precedence constraints and synchronization delays that are characteristic of a pipeline communication between maps (producers) and reduces (consumers). We here address the challenge of designing analytical models to estimate the performance of MapReduce workloads, notably Hadoop workloads, focusing particularly on the intra-job pipeline parallelism between map and reduce tasks belonging to the same job. We propose a hierarchical model that combines a precedence graph model and a queuing network model to capture the intra-job synchronization constraints. We first show how to build a precedence graph that represents the dependencies among multiple tasks of the same job. We then apply it jointly with an approximate Mean Value Analysis (aMVA) solution to predict mean job response time, throughput and resource utilization. We validate our solution against a queuing network simulator and a real setup in various scenarios, finding very close agreement in both cases. In particular, our model produces estimates of average job response time that deviate from measurements of a real setup by less than 15 %.

EVALUATING POTENTIAL FOR DATA ASSIMILATION IN A FLUX-TRANSPORT DYNAMO MODEL BY ASSESSING SENSITIVITY AND RESPONSE TO MERIDIONAL FLOW VARIATION

We estimate here a flux-transport dynamo model's response time to changes in meridional flow speed. Time-variation in meridional flow primarily determines the shape of a cycle in this class of dynamo models. In order to simultaneously predict the shape, amplitude and timing of a solar cycle by implementing an Ensemble Kalman Filter in the framework of Data Assimilation Research Testbed (DART), it is important to know the model's sensitivity to flow variation. Guided by observations we consider a smooth increase or decrease in meridional flow speed for a specified time (a few months to a few years), after which the flow speed comes back to the steady speed, and implement that time-varying meridional flow at different phases of solar cycle. We find that the model's response time to change in flow speed peaks at four to six months if the flow change lasts for one year. The longer the changed flow lasts, the longer the model takes to respond. Magnetic diffusivity has no influence in model's response to flow variation as long as the dynamo operates in the advection-dominated regime. Experiments with more complex flow variations indicate that the shape and amplitude of flow-perturbation have no influence in the estimate of model's response time.

Timing analysis of concurrent programs running on shared cache multi-cores

Memory accesses form an important source of timing unpredictability. Timing analysis of real-time embedded software thus requires bounding the time for memory accesses. Multiprocessing, a popular approach for performance enhancement, opens up the opportunity for concurrent execution. However due to contention for any shared memory by different processing cores, memory access behavior becomes more unpredictable, and hence harder to analyze. In this paper, we develop a timing analysis method for concurrent software running on multi-cores with a shared instruction cache. Communication across tasks is by message passing. Our method progressively improves the lifetime estimates of tasks that execute concurrently on multiple cores, in order to estimate potential conflicts in the shared cache. Possible conflicts arising from overlapping task lifetimes are accounted for in the hit-miss classification of accesses to the shared cache, to provide safe execution time bounds. We show that our method produces lower worst-case response time (WCRT) estimates than existing shared-cache analysis on a real-world embedded application. Furthermore, we also exploit instruction cache locking to improve WCRT. By locking some beneficial memory blocks into L1 cache, the WCET of the tasks and L2 cache conflicts are reduced, resulting in better WCRT. Experiments demonstrate that significant WCRT reduction is achieved through cache locking.

The influence of different Stop-signal response time estimation procedures on behavior–behavior and brain–behavior correlations

The fundamental cognitive-control function of inhibitory control over motor behavior has been extensively investigated using the Stop-signal task. The critical behavioral parameter describing stopping efficacy is the Stop-signal response time (SSRT), and correlations with estimates of this parameter are commonly used to establish that other variables (e.g., other behavioral measures or brain activity measures) are closely related to inhibitory motor control. Recently, however, it has been argued that SSRT estimates can be strongly distorted if participants strategically slow down their responses over the course of the experiment, resulting in the SSRT no longer reliably representing response-inhibition efficacy. Here, we performed new analyses on behavioral and functional data from an fMRI version of the Stop-signal task to gauge the consequences of using different SSRT estimation approaches that are differentially prone to the influence of strategic response slowing. The results indicate that the SSRT estimation approach can dramatically change behavior–behavior correlations. Specifically, a correlation between the SSRT and Go-trial accuracy that was highly significant with one estimation approach, virtually disappeared for the other. Additional analyses indeed supported that this effect was related to strategic response slowing. Concerning brain–behavior correlations, only the left anterior insula was found to be significantly correlated with the SSRT within the set of areas tested here. Interestingly, this brain–behavior correlation differed little for the different SSRT-estimation procedures. In sum, the current results highlight that different SSRT-estimation procedures can strongly influence the distribution of SSRT values across subjects, which in turn can ramify into correlational analyses with other parameters.

Long term change of nutrient concentrations of rivers discharging in European seas

Cases of severe eutrophication are still observed in European surface waters even though tough regulation has been in place since the beginning of the 1990s to control nutrient losses and inputs in the environment. The purpose of this paper is to evaluate the evolution since 1991 of the quality of the water entering European seas in terms of the concentration of major nutrients (nitrogen and phosphorus), and to analyze the effectiveness of implemented national/international measures and EU legislation in reducing water nutrient pollution. Despite the reduction in large portions of the European territory of agricultural nutrient applications and nutrient point source emissions, the impact on water quality is limited. It is shown using two large river basins that this lack of response for nitrogen, and nitrate in particular, between the reduction of the nitrogen surplus and the recovery of water quality is partly explained by the lag time due to transfer of nitrates in the unsaturated and saturated zones and storage in the soils and aquifers. In order to monitor efficiently the impact of policy implementation on water quality, the Nitrates Directive and the Urban Waste Water Treatment Directive in particular, it is recommended to use long term permanent monitoring stations to be able to separate the impact of climate variability from that of policy implementation. It is also recommended to investigate and develop harmonized methodologies for estimating the lag time in order to come up with realistic estimates of response time of water bodies due to the implementation of measures.

First Spike Latency Code for Interaural Phase Difference Discrimination in the Guinea Pig Inferior Colliculus

First spike latency has been suggested as a source of the information required for fast discrimination tasks. However, the accuracy of such a mechanism has not been analyzed rigorously. Here, we investigate the utility of first spike latency for encoding information about the location of a sound source, based on the responses of inferior colliculus (IC) neurons in the guinea pig to interaural phase differences (IPDs). First spike latencies of many cells in the guinea pig IC show unimodal tuning to stimulus IPD. We investigated the discrimination accuracy of a simple latency code that estimates stimulus IPD from the preferred IPD of the single cell that fired first. Surprisingly, despite being based on only a single spike, the accuracy of the latency code is comparable to that of a conventional rate code computed over the entire response. We show that spontaneous firing limits the capacity of the latency code to accumulate information from large neural populations. This detrimental effect can be overcome by generalizing the latency code to estimate the stimulus IPD from the preferred IPDs of the population of cells that fired the first n spikes. In addition, we show that a good estimate of the neural response time to the stimulus, which can be obtained from the responses of the cells whose response latency is invariant to stimulus identity, limits the detrimental effect of spontaneous firing. Thus, a latency code may provide great improvement in response speed at a small cost to the accuracy of the decision.

Measuring Client-Perceived Pageview Response Time of Internet Services

As e-commerce services are exponentially growing, businesses need quantitative estimates of client-perceived response times to continuously improve the quality of their services. Current server-side nonintrusive measurement techniques are limited to nonsecured HTTP traffic. In this paper, we present the design and evaluation a monitor, namely sMonitor, which is able to measure client-perceived response times for both HTTP and HTTPS traffic. At the heart of sMonitor is a novel size-based analysis method that parses live packets to delimit different webpages and to infer their response times. The method is based on the observation that most HTTP(S)-compatible browsers send significantly larger requests for container objects than those for embedded objects. sMonitor is designed to operate accurately in the presence of complicated browser behaviors, such as parallel downloading of multiple webpages and HTTP pipelining, as well as packet losses and delays. It requires only to passively collect network traffic in and out of the monitored secured services. We conduct comprehensive experiments across a wide range of operating conditions using live secured Internet services, on the PlanetLab, and on controlled networks. The experimental results demonstrate that sMonitor is able to control the estimation error within 6.7 percent, in comparison with the actual measured time at the client side.

Effect of type 2 diabetes on the dynamic response characteristics of leg vascular conductance during exercise

In this study we tested the hypothesis that type 2 diabetes impairs the dynamic response of leg vascular conductance (LVC) during exercise. LVC (leg blood flow/mean arterial pressure) responses were studied during intermittent contractions of the calf muscle in subjects with type 2 diabetes (n = 9), heavy controls (n = 10) and lean controls (n = 8) using a biexponential function and an estimate of the mean response time (MRT). The time constant of the second phase of LVC was significantly greater in type 2 diabetes (66.4 ± 29.2 s) than the heavy (22.2 ± 13.4 s) and lean (21.8 ± 9.3 s) controls, resulting in a significantly greater MRT in the diabetic group (median [IQR] = 30.7 [24.6-46.5] s versus 16.3 [4.3-23.2] s and 18.4 [13.7-19.3] s). These data support the hypothesis and suggest that a slowed hyperaemic response in the exercising limb might contribute to exercise intolerance in diabetic subjects.

Accuracy in temperature sensor response time estimation for new nuclear reactor designs

One method for measuring the response time of temperature sensors, the plunge test, verifies that the sensor has a suitable response time in the laboratory before installation. However, plunge test results cannot be extrapolated to the response time in an operating plant because response time is affected by multiple factors such as the ratio of internal heat-transfer resistance to the surface heat-transfer resistance, or Biot Modulus. The estimation method presented here can be used to extrapolate laboratory response-time measurements to determine sensor response time in another medium, in different test conditions, or in actual applications such as a nuclear power plant. However, experiments conducted at Oak Ridge National Laboratory have confirmed that the effect of temperature on sensor response time cannot be estimated confidently in an operating environment. Therefore, the loop current step response (LCSR) method was developed to measure the actual in-service response time of nuclear plant temperature sensors.

Nonindependent and nonstationary response times in stopping and stepping saccade tasks

Saccade stop signal and target step tasks are used to investigate the mechanisms of cognitive control. Performance of these tasks can be explained as the outcome of a race between stochastic go and stop processes. The race model analyses assume that response times (RTs) measured throughout an experimental session are independent samples from stationary stochastic processes. This article demonstrates that RTs are neither independent nor stationary for humans and monkeys performing saccade stopping and target-step tasks. We investigate the consequences that this has on analyses of these data. Nonindependent and nonstationary RTs artificially flatten inhibition functions and account for some of the systematic differences in RTs following different types of trials. However, nonindependent and nonstationary RTs do not bias the estimation of the stop signal RT. These results demonstrate the robustness of the race model to some aspects of nonindependence and nonstationarity and point to useful extensions of the model.

Population density and racial differences in the performance of emergency medical services

This paper analyzes the existence and scope of possible racial differences/disparities in the provision of emergency medical services (EMS) response capability (time from dispatch to arrival at the scene and level of training of the responding team) using data on approximately 120,000 cardiac incidents in the state of Mississippi during 1995–2004. The conceptual framework and empirical analysis focus on the likely effects of population density on the efficient production of EMS as a local public good subject to congestion, and on the need to control adequately for population density to avoid bias in testing for racial differences. Models that control for aggregate population density at the county-level indicate “reverse” disparities: faster estimated response times for African-Americans than for whites. When a refined county-level measure of population density is used that incorporates differences in African-American and white population density by Census tract, the reverse disparity in response times disappears. There also is little or no evidence of race-related differences in the certification level of EMS responders. However, there is evidence that, controlling for response time, African-Americans on average were significantly more likely to be deceased than whites upon EMS arrival at the scene. The overall results are germane to the debate over the scope of conditioning variables that should be included when testing for racial disparities in health care.

Liquid crystal display response time estimation for medical applications

Accurate characterization of diagnosis instruments is crucial in medical applications such as radiology and clinical neurosciences. While classical CRT medical displays have been replaced almost exclusively with liquid crystal devices (LCDs), the assessment of their temporal properties (response times) is still largely based on heuristic methods, which have not been evaluated thoroughly yet. The authors introduce a novel approach and show that it improves the accuracy and reliability compared to the common heuristic recommended by ISO 9241-305 substantially for a wide range of settings. The approach is based on disentangling the signal from the modulatory backlight through division (division approach). They evaluated this method in two different ways: First, they applied both methods to luminance transition measurements of different LCD monitors. Second, they simulated LCD luminance transitions by modeling the LCD optical responses according to a physical liquid crystal director orientation model. The simulated data were generated for four different response times, each with four different backlight modulation frequencies. Both the novel and the ISO convolution method were applied to the data. Application of the methods to the simulated data shows a bias of up to 46% for the ISO approach, while the novel division approach is biased at most 2%. In accordance with the simulations, estimates for real measurements show differences in the two approaches of more than 200% for some LCD panels. The division approach is robust against periodic backlight fluctuations and can reliably estimate even very short response times or small transitions. Unlike the established method, it meets the accuracy requirements of medical applications. In contrast, the popular convolution approach for estimating response times is prone to misestimations of time by several orders of magnitude and tend to further worsen as advances in LCD technology lead to shorter response times.

Relation between derived-band auditory brainstem response latencies and behavioral frequency selectivity

Derived-band click-evoked auditory brainstem responses (ABRs) were obtained for normal-hearing (NH) and sensorineurally hearing-impaired (HI) listeners. The latencies extracted from these responses, as a function of derived-band center frequency and click level, served as objective estimates of cochlear response times. For the same listeners, auditory-filter bandwidths at 2 kHz were estimated using a behavioral notched-noise masking paradigm. Generally, shorter derived-band latencies were observed for the HI than for the NH listeners. Only at low click sensation levels, prolonged latencies were obtained for some of the HI listeners. The behavioral auditory-filter bandwidths accounted for the across-listener variability in the ABR latencies: Cochlear response time decreased with increasing filter bandwidth, consistent with linear-system theory. The results link cochlear response time and frequency selectivity in human listeners and offer a window to better understand how hearing impairment affects the spatiotemporal cochlear response pattern.

Estimation of cochlear response times using lateralization of frequency-mismatched tones

Behavioral and objective estimates of cochlear response times (CRTs) and traveling-wave (TW) velocity were compared for three normal-hearing listeners. Differences between frequency-specific CRTs were estimated via lateralization of pulsed tones that were interaurally mismatched in frequency, similar to a paradigm proposed by Zerlin [(1969). J. Acoust. Soc. Am. 46, 1011-1015]. In addition, derived-band auditory brainstem responses were obtained as a function of derived-band center frequency. The latencies extracted from these responses served as objective estimates of CRTs. Estimates of TW velocity were calculated from the obtained CRTs. The correspondence between behavioral and objective estimates of CRT and TW velocity was examined. For frequencies up to 1.5 kHz, the behavioral method yielded reproducible results, which were consistent with the objective estimates. For higher frequencies, CRT differences could not be estimated with the behavioral method due to limitations of the lateralization paradigm. The method might be useful for studying the spatiotemporal cochlear response pattern in human listeners.

Pace: An Alternative Measure of Student Question Response Time

In some professions, speed and accuracy are as important as acquired requisite knowledge and skills. The availability of computer-based testing now facilitates examination of these two important aspects of student performance. We found that student response times in a conventional non-speeded multiple-choice test, at both the global and individual question levels, closely approximated lognormal distributions. We propose a new measure, pace, which is derived from the survival function of these distributions for analysis of individual person response times. These pace estimates could be used both to rank and compare students; pace also performed maximally compared to other parameterizations in generalizability and dependability studies. While pace was very weakly related to person ability, there was no detectable relationship to question parameters of shift, natural logarithmic mean, or natural logarithmic standard deviation. That is, pace was a person-dependent, question-independent measure. Pace measurements were also successfully used as covariates in models for estimation of person response time to specified questions and person accuracy in response to specified questions. Thus, the analysis of pace can contribute significantly to comprehensive evaluation of student performance in both the speed and ability domains and is a requisite to best practice in testing and assessment.