The impact of predictive inaccuracies on execution scheduling

This paper investigates the underlying impact of predictive inaccuracies on execution scheduling, with particular reference to execution time predictions. This study is conducted from two perspectives: from that of job selection and from that of resource allocation, both of which are fundamental components in execution scheduling. A new performance metric, termed the degree of misperception, is introduced to express the probability that the predicted execution times of jobs display different ordering characteristics from their real execution times due to inaccurate prediction. Specific formulae are developed to calculate the degree of misperception in both job selection and resource allocation scenarios. The parameters which influence the degree of misperception are also extensively investigated. The results presented in this paper are of significant benefit to scheduling approaches that take into account predictive data; the results are also of importance to the application of these scheduling techniques to real-world high-performance systems.     

基于可靠性的云工作流调度策略

经过对已有云工作流调度算法中可靠性问题进行分析研究,针对一些算法在任务调度过程中只考虑提高整个工作流的可靠性而牺牲了时间或增加花费的问题,结合云计算的特点,提出一种基于可靠性的工作流调度策略。该策略结合了工作流中任务的可靠性,充分考虑任务的优先顺序并结合复制的思想,在减少传输过程失败率的同时降低传输时间,使整个工作流在降低完成时间的同时,提高整体可靠性。通过实验和分析表明,通过该策略云工作流在不同任务数和通信运算比（CCR）的可靠性比异态最早结束时间算法（HEFT）算法及其改进算法--SHEFTEX都有所提升,完成时间比HEFT算法有所减少。     

Hadoop平台下改进的推测任务调度算法

研究对比Hadoop平台下默认的推测任务调度算法和异构环境下LATE调度算法的优势和不足,提出了一种基于Hadoop集群的改进的推测任务调度算法.该算法以节点历史信息对Reduce任务各阶段比例进行动态调整和更新,并对任务实时处理速率进行局部平滑处理来提高预估任务剩余完成时间的准确性,最后采用MCP模型对备份任务有效性进行验证.通过实验结果分析可知:该算法能够有效提升备份任务成功率,减少作业完成时间.     

Efficient consolidation-aware VCPU scheduling on multicore virtualization platform

Multicore processors are widely used in today’s computer systems. Multicore virtualization technology provides an elastic solution to more efficiently utilize the multicore system. However, the Lock Holder Preemption (LHP) problem in the virtualized multicore systems causes significant CPU cycles wastes, which hurt virtual machine (VM) performance and reduces response latency. The system consolidates more VMs, the LHP problem becomes worse. In this paper, we propose an efficient consolidation-aware vCPU (CVS) scheduling scheme on multicore virtualization platform. Based on vCPU over-commitment rate, the CVS scheduling scheme adaptively selects one algorithm among three vCPU scheduling algorithms: co-scheduling, yield-to-head, and yield-to-tail based on the vCPU over-commitment rate because the actions of vCPU scheduling are split into many single steps such as scheduling vCPUs simultaneously or inserting one vCPU into the run-queue from the head or tail. The CVS scheme can effectively improve VM performance in the low, middle, and high VM consolidation scenarios. Using real-life parallel benchmarks, our experimental results show that the proposed CVS scheme improves the overall system performance while the optimization overhead remains low.     

Optimal feedback scheduling of model predictive controllers

Model predictive control （MPC） could not be reliably applied to real-time control systems because its computation time is not well defined. Implemented as anytime algorithm, MPC task allows computation time to be traded for control performance, thus obtaining the predictability in time. Optimal feedback scheduling （FS-CBS） of a set of MPC tasks is presented to maximize the global control performance subject to limited processor time. Each MPC task is assigned with a constant bandwidth server （CBS）, whose reserved processor time is adjusted dynamically. The constraints in the FS- CBS guarantee scheduler of the total task set and stability of each component. The FS-CBS is shown robust against the variation of execution time of MPC tasks at runtime. Simulation results illustrate its effectiveness.     

M-WiMAX中马尔可夫预测的PF调度算法研究

在M-WiMAX系统中为了获得多用户分集增益,更好地发挥自适应调制编码（AMC）的作用,在改进比例公平算法（PF）基础上设计出一个预测的PF分组调度算法（Pre-PF）来调度系统中的非实时业务。PF调度算法在调度判决表达式中用调度时刻瞬时速率作为参数,这样影响调度的稳定性和整体性能,尤其在移动无线通信中更加明显。提出用马尔可夫预测模型来计算下一个调度时刻的瞬时速率,用它来平滑当前瞬时速率,减小移动无线信道瞬时速率对调度性能的影响。平滑后的瞬时速率能更好地代表移动信道传输速率的变化趋势。仿真结果表明,该调度算法保证了非实时业务流的服务质量,提高了系统的平均吞吐量、稳定性和公平性。     

云平台下基于截止时间的自适应调度策略

针对在共享集群中进行任务调度时,无法兼顾任务的响应速度与任务完成时间的问题,提出一种基于截止时间的自适应调度算法。该算法以用户提交的截止时间为依据,根据任务的执行进度自适应地分配适当的计算资源。不同于传统调度方式里由用户提交固定资源参数,该算法在资源约束的情况下会对优先级高的任务进行抢占式调度以保证服务质量（QoS）,并在抢占过程结束后额外分配资源补偿被抢占的任务。在Spark平台进行的任务调度实验结果显示,与另一种资源协调者（YARN）框架下的调度算法相比,所提算法能严格地控制短任务的响应速度,并使长作业的任务完成时间缩短35%。     

Hadoop平台下基于截止时间限制的动态调度算法的研究

为了满足有截止时间限制的MapReduce作业的需求,提出一种基于截止时间限制的动态调度算法（DCDS）。该算法实时监控作业运行状况,并对作业运行时间进行动态估算,从而确定作业优先级;对于时间紧迫的作业,可通过抢占策略来保证在用户要求的截止时间内完成。实验结果表明,与Hadoop平台现有的调度算法相比,该算法不仅能满足作业截止时间的要求,也提高了系统资源的利用率和吞吐量。     

IaaS公有云平台调度模型研究

抽象出IaaS公有云平台的服务模型,基于排队论对平台服务模式、队列长度、调度服务器设置等进行了优化分析。在此基础上提出一种基于IaaS平台需求向量的调度模型,根据需求与可用资源的匹配度从平台管理的物理机集合中筛选出可用的宿主机,若一次性无法找到符合要求的宿主机,平台调度算法结合虚拟机迁移操作,对物理资源进行重新分配,在实现平台资源利用率最大化的同时,保障了平台的可用性。将该算法应用在自主研发的云计算平台上,实验结果验证了该算法的可行性。     

无线传感器网络节点调度优化分簇算法研究*

节点调度优化分族算法根据网络节点部署的特点,在不影响网络服务质量的情况下,通过冗余节点调度优化分族的方法能有效减少网络中工作节点的数量,从而节约网络平均每一轮的网络能耗.仿真结果表明,与未考虑冗余节点状态调度的算法相比,节点调度优化分簇算法有效提高了网络能量利用率,延长了网络寿命.     

基于SaaS的弹性云平台优化调度策略设计

云计算平台利用虚拟化技术使软件应用变得更有效率的同时, 也给资源管理和服务调度带来了挑战。在研究了软件服务（SaaS）与基础设施服务（IaaS）调度的区别基础上, 重点考虑SaaS层的资源调度, 提出基于随机理论的调度模型, 把该层调度描述成一种多目标的优化问题。除了服务质量的要求, 还考虑了弹性这一云服务的重要特性, 并提供了任务调度与弹性服务副本的匹配策略。实验表明本调度机制的设计优化了云平台的整体性能, 达到了较好的负载均衡与资源利用率。     

基于随机模型的云平台资源调度策略设计

针对云计算资源管理的实际需求,提出一种基于随机模型的云平台调度策略,设计合理高效的资源调度算法,解决传统代数模型请求丢失率高以及其他随机模型负载均衡指标性能较差的问题,从而在服务性能和执行效率的基础上保证服务器的资源负载,使云平台处于相对稳定的状态。在实验环境中的验证结果表明,该调度策略能够优化虚拟资源的使用效率和服务响应时间,同时能够达到较好的负载均衡并降低运营成本。     

Hadoop平台公平调度算法研究与优化

Hadoop Map Reduce框架的公平调度算法以统一的固定配置文件管理计算节点上计算槽的数量,这不能保障集群负载均衡,亦不能满足不同用户的资源需求。针对公平调度算法配置方式的不足,提出一种动态反馈的调度算法。该算法结合公平调度算法预先分配的特性,能够对计算节点上的计算槽进行动态调整。实验结果表明,基于动态反馈的改进算法有效地提高了集群的执行效率。     

Hadoop平台的多队列作业调度优化方法研究

对Hadoop平台的作业调度算法进行了研究, 提出了支持作业类型区分的多队列调度优化算法。优化算法支持根据节点当前的负载情况分配不同类型的作业, 以提高节点的资源利用率; 允许作业队列的资源在闲置时被其他作业队列占用; 在原作业队列需要时可以被即时回收, 即回收过程支持任务抢占; 采用共享队列列表和非共享队列列表的逻辑划分来防止乒乓效应。Hadoop平台的性能测试结果表明, 优化算法相比系统默认算法在作业调度的执行效率、执行平稳性等方面都有了显著的提升。     

Production scheduling of parallel machines with model predictive control

This paper considers the problem of optimizing on-line the production scheduling of a multiple-line production plant composed of parallel equivalent machines which can operate at different speeds corresponding to different energy demands. The transportation lines may differ in length and the energy required to move the part to be processed along them is suitably considered in the computation of the overall energy consumption. The optimal control actions are recursively computed with Model Predictive Control aiming to limit the total energy consumption and maximize the overall production. Simulation results are reported to witness the potentialities of the approach in different scenarios.     

无线传感器节点定位功能测试平台

建模测试对无线传感器网络的研究以及相关商用产品的开发起着至关重要的作用.描述了用于监控无线传感器节点定位功能的测试平台的工作过程,构建了基于ARM嵌入式系统的无线传感器网络.采用相对定位方法,确定网络内各节点的相对位置,降低误差,以符合精度要求.分析了在μC/OS-Ⅱ操作系统的调度下,传感器节点利用GPS模块进行相对定位的流程,并对网络状态的显示界面作了相应介绍.利用该平台对传感器网络进行测试,所得定位数据符合实际情况,获得了预期的结果.     

Visual sensor network lifetime maximization by prioritized scheduling of nodes

Visual surveillance of a designated air space can be achieved by a randomly distributed camera sensor network spread over a large area. The location and field of view of each battery operated sensor, after a calibration phase, will be known to a central processing node. To increase the lifetime of the network, the density of distributed sensors could be such that a subset of sensors can cover the required air space. As a sensor dies another sensor should be selected to compensate for the dead one and reestablish the complete coverage. This process should be continued until complete coverage is not achievable by the existing sensors. Thereafter, a graceful degradation of the coverage is desirable.The goal is to elongate the lifetime of the network while maintaining a maximum possible coverage of the designated air space. Since the selection of a subset of sensors for complete coverage of the target area is an NP-complete problem, we present a number of heuristics for this case. The proposed methods are categorized in two groups. In one category, the sensors are prioritized based on their visual and communicative properties and the selection is performed according to the prioritizing function. In the other group, we propose traditional evolutionary and swarm intelligence algorithms. The performance of the proposed methods is evaluated through extensive simulations.     

基于隶属度函数的车间协同调度方法*

快速发现制造瓶颈是实现车间调度的有效方法。借鉴面向对象的设计思想以及基于模糊数学的最大隶属度原则,提出面向产能瓶颈单元的协同调度,进一步提出产能协同调度二维表示方法,将各类资源归一化,研究并建立了产能协同调度模型。通过实证,运用面向瓶颈的优化调度方法,解决了如何在有限资源的情况下实现复杂制造系统最优化调度的问题。     

Integration of process planning and scheduling functions

Automated process planning (APP) has received a considerable amount of attention in recent years. This has mainly been due to APP allowing for the bridging of the gap between design and manufacturing, and for its automation of the related tedious human task. While there have been several efforts in developing interfaces between computer-aided design (CAD) and APP, little attention has been given to the integration of APP and scheduling, which is likely to be of even higher importance. This paper addresses the basic issues involved in the integration of the two functions, demonstrates a methodology and the potential benefits of the integrated approach by an example, and presents directions for future research.