A major obstacle towards the adoption of multi-core platforms for real-time systems is given by the difficulties in characterizing the interference due to memory contention. The simple fact that multiple cores may sim...
详细信息
ISBN:
(纸本)9781450335911
A major obstacle towards the adoption of multi-core platforms for real-time systems is given by the difficulties in characterizing the interference due to memory contention. The simple fact that multiple cores may simultaneously access shared memory and communication resources introduces a significant pessimism in the timing and schedulability analysis. To counter this problem, predictable execution models have been proposed splitting task executions into two consecutive phases: a memory phase in which the required instruction and data are pre-fetched to local memory (M-phase), and an execution phase in which the task is executed with no memory contention (C-phase). Decoupling memory and execution phases not only simplifies the timing analysis, but it also allows a more efficient (and predictable) pipelining of memory and execution phases through proper co-scheduling algorithms. In this paper, we take a further step towards the design of smart co-scheduling algorithms for sporadic real-time tasks complying with the M/C (memory-computation) model. We provide a theoretical framework that aims at tightly characterizing the schedulability improvement obtainable with the adopted M/C task model on a single-core systems. We identify a tight critical instant for M/C tasks scheduled with fixed priority, providing an exact response-time analysis with pseudo-polynomial complexity. We show in our experiments that a significant schedulability improvement may be obtained with respect to classic execution models, placing an important building block towards the design of more efficient partitioned multi-core systems.
The contribution deals with local search based optimization of schedules that are modelled by Petri nets (PNs). PN modelling of scheduling problems in manufacturing domain is addressed and a generalization of neighbou...
详细信息
With voltage frequency islanding (VFI) we are able to change the voltage and frequency of the processors to save power within a certain performance constraint. In this work we want to maximize potential power savings ...
详细信息
ISBN:
(纸本)9781509001729
With voltage frequency islanding (VFI) we are able to change the voltage and frequency of the processors to save power within a certain performance constraint. In this work we want to maximize potential power savings and minimize time penalty by means of efficient task scheduling in VFIs. A new scheduling algorithm is presented which schedules the tasks onto the processing cores in the VFI by ranking them based on the computation load and communication load among other tasks. The mapping of the tasks is relevant to the processing speed of the cores and the simulation results show improvements compared to the default task scheduler of Linux.
With growing data volumes generated and stored across geo-distributed datacenters, it is becoming increasingly inefficient to aggregate all data required for computation at a single datacenter. Instead, a recent trend...
详细信息
ISBN:
(纸本)9781450336512
With growing data volumes generated and stored across geo-distributed datacenters, it is becoming increasingly inefficient to aggregate all data required for computation at a single datacenter. Instead, a recent trend is to distribute computation to take advantage of data locality, thus reducing the resource (e.g., bandwidth) costs while improving performance. In this trend, new challenges are emerging in job scheduling, which requires coordination among the datacenters as each job runs across geo-distributed sites. In this paper, we propose novel job scheduling algorithms that coordinate job scheduling across datacenters with low overhead, while achieving near-optimal performance. Our extensive simulation study with realistic job traces shows that the proposed scheduling algorithms result in up to 50% improvement in average job completion time over the Shortest Remaining Processing Time (SRPT) based approaches.
Fairness and data locality are often in conflict in Hadoop job scheduling. During scheduling, it is not always possible for data locality to be achieved for all jobs or for fairness to be attained for all users. Achie...
详细信息
ISBN:
(纸本)9781509001446
Fairness and data locality are often in conflict in Hadoop job scheduling. During scheduling, it is not always possible for data locality to be achieved for all jobs or for fairness to be attained for all users. Achieving pure fairness may compromise the data locality of the jobs which will negatively affect performances, and vice-versa. For example, a scheduler may opt to sacrifice performance by scheduling tasks to non-data local nodes. Alternatively, a scheduler may choose to sacrifice fairness by giving up an available slot and wait for a data-local node. The Dynamic Task Splitting Scheduler (DTSS) is proposed to mitigate the tradeoffs between fairness and data locality during job scheduling. DTSS does so by dynamically splitting a task and executing the split task immediately, on a non-data-local node, to improve the fairness. Analysis and experiments results show that it is possible to improve both fairness and the performance by adjusting the proportion of the task split. DTSS is shown to improve the makespan of different users in a cluster by 2% to 11% as compared to delay scheduling under the situation where it is difficult to obtain data-local nodes on a cluster. Lastly, experiments show that DTSS is not a suitable scheduler under conditions where jobs are able to obtain data-local nodes easily.
Multicore microcontrollers are rapidly penetrating the real-time systems market with a promise of increased processing throughput and lower energy consumption compared to traditional single-core processors. Consequent...
详细信息
ISBN:
(纸本)9781479959273
Multicore microcontrollers are rapidly penetrating the real-time systems market with a promise of increased processing throughput and lower energy consumption compared to traditional single-core processors. Consequently, it sparked a wave of research in Soft Real-Time (SRT) scheduling algorithms for multicore processors producing many different scheduling algorithms that are good at harnessing the added power of the multicore processors but lack the required determinism to be used in Hard Real-Time (HRT) systems. We present the Multicore Priority Ceiling Protocol scheduling algorithm for HRT systems along with its schedulability test for HRT systems running on a single instance of a modified RTOS. We also developed a tool that uses our schedulability test to solve the task partitioning problem using a Particle Swarm Optimization.
Current evaluations of DoD tactical networking systems have left the impression of sub-optimal performance without being able to provide either a clear vision of the limits on the performance that could realistically ...
详细信息
ISBN:
(纸本)9781509000739
Current evaluations of DoD tactical networking systems have left the impression of sub-optimal performance without being able to provide either a clear vision of the limits on the performance that could realistically be attained or which layers of the network algorithm are responsible for the suboptimal results. Recent advances in computing the capacity of a large MANET allow us to obtain a practical benchmark for MANET capacity performance evaluation [1]. In this paper, we quantify the impact of commonly used algorithms at the routing and the scheduling layers on the overall network throughput and compare their individual effects on overall network throughput performance. We consider the routing and the scheduling layers separately since practical MANET implementations are likely to use a layered architecture even though the joint routing and scheduling algorithm is known to be optimal. Our data shows that a good scheduling algorithm can provide potentially four times the throughput improvement of a good routing algorithm when inter-user interference conforms to an 802.11 type model.
Traffic signals are essential to provide safe driving that allows all traffic flows to share road intersection. However, they decrease the traffic flow fluency because of the queuing delay at each road intersection. I...
详细信息
ISBN:
(纸本)9783319272931;9783319272924
Traffic signals are essential to provide safe driving that allows all traffic flows to share road intersection. However, they decrease the traffic flow fluency because of the queuing delay at each road intersection. In order to improve the traffic efficiency all over the road network, Intelligent Traffic Light scheduling (ITLS) algorithm has been proposed. In this work, we introduce an ITLS algorithm based on Genetic Algorithm (GA) merging with Machine Learning (ML) algorithm. This algorithm schedules the time phases of each traffic light according to each real-time traffic flow that intends to cross the road intersection, whilst considering next time phases of traffic flow at each intersection by ML. In order to get each next time phases of traffic flow, we use Linear Regression (LR) algorithm as ML algorithm. The introduced algorithm aims to increase traffic fluency by decreasing the total waiting delay of all traveling vehicles at each road intersection in the road network. We compare the performance of our algorithm with the unimproved one for different simulated data. Results shows that, our algorithm increases the traffic fluency and decreases the waiting delay by 21.5 % compared with the unimproved one.
In this paper five scheduling algorithms were investigated and their performance was evaluated in terms of Fairness Index, Peak Throughput, Average Throughput and Edge Cell User Throughput. A system level MATLAB simul...
详细信息
ISBN:
(纸本)9783319076744;9783319076737
In this paper five scheduling algorithms were investigated and their performance was evaluated in terms of Fairness Index, Peak Throughput, Average Throughput and Edge Cell User Throughput. A system level MATLAB simulator was used. The simulation takes into account different types of traffic for several mobility scenarios and propagation channel models. Results indicate that the scheduling algorithms showed some quality in certain parameter of evaluation but lack in other terms. While some scheduling algorithm take the moderate path but still be lacking especially in Edge Cell User Throughput necessitating the use of Relays or femtocells.
Mixed-criticality systems emerged with the aim of reconciling safety requirements and efficient use of multi processor or uniprocessor platforms. On multi processors, recent works on mixed criticality have produced im...
详细信息
ISBN:
(纸本)9781467379298
Mixed-criticality systems emerged with the aim of reconciling safety requirements and efficient use of multi processor or uniprocessor platforms. On multi processors, recent works on mixed criticality have produced impressive results in terms of speed up factor. But these solutions, based on Pfair like scheduling algorithms, entail too many preemptions and migrations to be effectively used in real systems. As RUN is an optimal scheduling algorithm that is known to limit this problem, we propose MxC-RUN, an adaptation of RUN to mixed -criticality systems. We redefine RUN's primal servers as modal servers that allocate the overestimated time budget of their higher criticality tasks to execute lower criticality ones. These servers can be handled by RUN without any modification and preserve its performances in terms of preemptions and migrations. MxC-RUN earns a speed up factor smaller than other multi processors EDF-based mixed-criticality scheduling algorithms.
暂无评论