Number Of Concurrent Tasks Research Articles

Joint optimization of task scheduling and computing resource allocation for VR video services in 5G‐advanced networks

AbstractWith the development of 5G networks and related hardware equipment, the demand for various VR video services including panoramic video live broadcast and immersive game is increasing day by day. Due to the explosive growth of data volume and the trend toward lightweight VR equipment, traditional local computing architecture faces problems of insufficient computing power and high cost. As an emerging computing paradigm, the remote computing architecture of cloud‐edge collaboration frees users from cumbersome local restrictions. However, edge computing resources are often limited, and the remote cloud has inconsistent network latency. Therefore, how to adaptively schedule and allocate computing resources is the key to determining VR service quality. Based on an in‐depth analysis of the characteristics of VR video services and network environments, this article first builds a cloud‐edge‐network coorperation computing and transmission architecture for VR video services. This architecture can adaptively offload concurrent computing tasks and allocate computing power to them in VR scenes based on network conditions provided by the 5G‐Advanced networks via network information exposure interface. This paper proposes a general optimization problem to maximize user experience. And a single‐slot delay optimization model was constructed. Combining convex optimization theory and deep learning algorithms, this paper designs a task scheduling and computing power allocation strategy based on deep reinforcement learning. The policy gradually learns optimal task scheduling and resource allocation decisions to maximize delay‐related rewards in the long run. This article further introduces the graph neural network under the attention mechanism to solve the problem of changes in the number of concurrent tasks. Finally, through Cloudsim simulation, this article verifies the superiority of task scheduling and computing power allocation strategies based on deep reinforcement learning compared to other common strategies in terms of resource utilization, client latency, scalability and decision time.

Self-Learning Threshold-Based Load Balancing

We consider a large-scale service system where incoming tasks have to be instantaneously dispatched to one out of many parallel server pools. The user-perceived performance degrades with the number of concurrent tasks and the dispatcher aims at maximizing the overall quality of service by balancing the load through a simple threshold policy. We demonstrate that such a policy is optimal on the fluid and diffusion scales, while only involving a small communication overhead, which is crucial for large-scale deployments. In order to set the threshold optimally, it is important, however, to learn the load of the system, which may be unknown. For that purpose, we design a control rule for tuning the threshold in an online manner. We derive conditions that guarantee that this adaptive threshold settles at the optimal value, along with estimates for the time until this happens. In addition, we provide numerical experiments that support the theoretical results and further indicate that our policy copes effectively with time-varying demand patterns. Summary of Contribution: Data centers and cloud computing platforms are the digital factories of the world, and managing resources and workloads in these systems involves operations research challenges of an unprecedented scale. Due to the massive size, complex dynamics, and wide range of time scales, the design and implementation of optimal resource-allocation strategies is prohibitively demanding from a computation and communication perspective. These resource-allocation strategies are essential for certain interactive applications, for which the available computing resources need to be distributed optimally among users in order to provide the best overall experienced performance. This is the subject of the present article, which considers the problem of distributing tasks among the various server pools of a large-scale service system, with the objective of optimizing the overall quality of service provided to users. A solution to this load-balancing problem cannot rely on maintaining complete state information at the gateway of the system, since this is computationally unfeasible, due to the magnitude and complexity of modern data centers and cloud computing platforms. Therefore, we examine a computationally light load-balancing algorithm that is yet asymptotically optimal in a regime where the size of the system approaches infinity. The analysis is based on a Markovian stochastic model, which is studied through fluid and diffusion limits in the aforementioned large-scale regime. The article analyzes the load-balancing algorithm theoretically and provides numerical experiments that support and extend the theoretical results.

Event-related potentials as indices of mental workload while using an in-vehicle information system

New in-vehicle information systems are now being commercialized. Despite the expected benefits, some concerns exist that they may overload drivers’ capacity and decrease performance. According to the multiple resource theory (Wickens, Hum Factors 50:449–455, https://doi.org/10.1518/001872008X288394, 2008), overload may occur at different stages of processing, that is, perceptual–central and/or response-related stages. Therefore, different measures may be needed to detect such specific demands. We explored the sensitivity of different mental workload measurements during the performance of an auditory task alone (single task) and in combination with a tracking task that was presented without (dual task) or, with a visual display (triple task). The demands associated with the number of concurrent tasks (single, dual and triple tasks), tracking speed (low, high, adjustable) and their interaction were analyzed. To account for different processing requirements, mental workload was assessed using subjective, behavioral (performance on the auditory task) and psychophysiological measurements (event-related potentials). 17 young adults participated in the study. The results showed that most measurements discriminated between the performances of one or more tasks, as well as between low and high speeds. However, only the subjective ratings and tracking task performance further discriminated between the dual- and triple-task conditions. Finally, ERPs (N1 and P3) were the only measure detecting increases in cognitive demands associated with higher requirements on processing speed combined with the addition of the display. Our results suggest that ERPs may provide complementary information to other traditional mental workload measures. Its applications in the evaluation and design of future systems should be investigated.

BOUNDS ON THE SCALABILITY OF BAG-OF-TASKS APPLICATIONS RUNNING ON MASTER-SLAVE PLATFORMS

Bag-of-Tasks applications are parallel applications composed of independent (i.e., embarrassingly parallel) tasks that do not communicate with each other, may depend upon one or more input files, and can be executed in any order. Each file may be input for more than one task. A common framework to execute BoT applications is the master-slave topology, in which the user machine is used to control the execution of tasks. In this scenario, a large number of concurrent tasks competing for resources (e.g., CPU and communication links) severely limits the scalability. In this paper we studied the scalability of BoT applications running on multi-node systems (e.g. clusters and grids) organized as master-slave platforms, considering two communications paradigms: multiplexed connections and efficient broadcast. We prove that the lowest bound possible on the isoefficiency function for master-slave platforms is achievable by those platforms that have an O(1) efficient broadcast primitive available. We also analyze the impact of output file contention in scalability, under different assumptions. Our study employs a set of simulation experiments that confirms and extends the theoretical results (e.g. by simulating TCP links).

Fluctuating Efforts and Interdependencies in Collaborative Work

We present a dynamical model of cooperative efforts comprised of concurrently performed, interrelated tasks. The model contains a stochastic component to account for temporal fluctuations both in task performance and in the effect of a given unit of work on the project as a whole. We show that as the number of concurrent tasks increases, so does the average completion time. Also, for fixed system size, the dynamics of individual project realizations can exhibit large deviations from the average when fluctuations increase past a certain threshold, causing long delays in completion times. These effects are in agreement with empirical observation. We also show that the negative effects of both large groups and long delays caused by fluctuations may be mitigated by arranging projects in a hierarchical or modular structure. Our model is applicable to any arrangement of interdependent tasks, providing an analytical prediction for the average completion time as well as a numerical threshold for the fluctuation strength beyond which long delays are likely. In conjunction with previous modeling techniques, it thus provides managers with a predictive tool to be used in the design of a project’s architecture.

Threads without the Pain

Much of today’s software deals with multiple concurrent tasks. Web browsers support multiple concurrent HTTP connections, graphical user interfaces deal with multiple windows and input devices, and Web and DNS servers handle concurrent connections or transactions from large numbers of clients. The number of concurrent tasks that needs to be handled increases while software grows more complex. Structuring concurrent software in a way that meets the increasing scalability requirements while remaining simple, structured, and safe enough to allow mortal programmers to construct ever-more complex systems is a major engineering challenge.

An Empirical Validation of Subjective Workload Ratings

A correlational design was used to regress NASA TLX subjective workload ratings onto several potential independent variables (i.e., the number of concurrent tasks, task combination, task resource demands, and flight experience) to determine task characteristics that influence pilot subjective workload ratings. A part task simulator was used to present up to six concurrent tasks, in different combinations, to 27 cadets at the USAF Academy. The results indicated the number of concurrent tasks had the largest impact on subjective workload ratings. In terms of multiple resource theory, spatial, verbal, and visual demands (in that order) contributed the most variance. The implication for theoreticians and designers is that the number of concurrent tasks, mental resource demands, and time constraints seem to be key contributors to subjective workload ratings.

The Prediction of Pilot Subjective Workload Ratings

This paper investigated potential variables that predict pilot subjective workload ratings. A correlational design was used to regress NASA TLX subjective workload ratings onto four independent variables (the number of concurrent tasks, task type, task combination, and flight experience). A pilot selection simulator was used to present five different flight tasks, in different combinations, to 15 cadets at the USAF Academy. The results indicated the number of concurrent tasks had the largest impact on subjective workload ratings, followed by subjects' flight experience. The data also showed that the effort dimension of the NASA TLX contributed the most variance to the overall weighted workload ratings. The implication for theoreticians and designers is that the number of concurrent tasks, experience, and perceived effort seem to be key contributor to subjective workload ratings.

Reliability analysis of a simple replicated and-fork/and-join graph

In this paper, we consider the problem of calculating the reliability of the replicated and-fork/and-join (RAFAJ) task graphs. An RAFAJ task graph is generated from an AFAJ task graph with vertex and arc replications. Due to the AFAJ constraint, the task graphs considered in our problem are more general than those considered in the network reliability problem, which is an NP-complete problem for general task graphs. In this paper, we propose an algorithm with time complexity O( n · 2 2 r ); where n is number of concurrent tasks, and r is the number of replications.