Energy-Aware Data Allocation and Task Scheduling on Heterogeneous Multiprocessor Systems With Time Constraints

Wang, Yan; Li, Kenli; Chen, Hao; He, Ligang; Li, Keqin

doi:10.1109/tetc.2014.2300632

Cited by 63 publications

(79 citation statements)

References 33 publications

(47 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Chetto [25] studied the scheduling for real-time jobs, which are executed on a uniprocessor system supplied by a renewable energy source. Wang et al [26] studied the energy-aware data allocation and task scheduling problem on multiprocessor system for real-time applications. Lin et al [27] studied the problem of scheduling co-design for reliability and energy by minimizing total energy while guaranteeing reliability constraints.…”

Section: Related Workmentioning

confidence: 99%

Efficient Energy Consumption Scheduling: Towards Effective Load Leveling

2017

View full text Add to dashboard Cite

Different agents in the smart grid infrastructure (e.g., households, buildings, communities) consume energy with their own appliances, which may have adjustable usage schedules over a day, a month, a season or even a year. One of the major objectives of the smart grid is to flatten the demand load of numerous agents (viz. consumers), such that the peak load can be avoided and power supply can feed the demand load at anytime on the grid. To this end, we propose two Energy Consumption Scheduling (ECS) problems for the appliances held by different agents at the demand side to effectively facilitate load leveling. Specifically, we mathematically model the ECS problems as Mixed-Integer Programming (MIP) problems using the data collected from different agents (e.g., their appliances' energy consumption in every time slot and the total number of required in-use time slots, specific preferences of the in-use time slots for their appliances). Furthermore, we propose a novel algorithm to efficiently and effectively solve the ECS problems with large-scale inputs (which are NP-hard). The experimental results demonstrate that our approach is significantly more efficient than standard benchmarks, such as CPLEX, while guaranteeing near-optimal outputs.

show abstract

Section: Related Workmentioning

confidence: 99%

Efficient Energy Consumption Scheduling: Towards Effective Load Leveling

2017

View full text Add to dashboard Cite

show abstract

“…In Section 5, we propose two heuristic algorithms to solve the Heterogeneous Data Allocation and Task Scheduling (HDATS) [8] problem. Related works are discussed in Section 2.…”

Section: Our Contributionsmentioning

confidence: 99%

“…These benchmarks are from DSP st one [16], and frequently used on multi-core systems [8]. These benchmarks are from DSP st one [16], and frequently used on multi-core systems [8].…”

Section: Experimental Apparatusmentioning

confidence: 99%

Data‐aware task scheduling on heterogeneous hybrid memory multiprocessor systems

Chen

Liu

et al. 2016

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

In this paper, we propose a method about task scheduling and data assignment on heterogeneous hybrid memory multiprocessor systems for real-time applications. In a heterogeneous hybrid memory multiprocessor system, an important problem is how to schedule real-time application tasks to processors and assign data to hybrid memories. The hybrid memory consists of dynamic random access memory and solid state drives when considering the performance of solid state drives into the scheduling policy. To solve this problem, we propose two heuristic algorithms called improvement greedy algorithm and the data assignment according to the task scheduling algorithm, which generate a near-optimal solution for real-time applications in polynomial time. We evaluate the performance of our algorithms by comparing them with a greedy algorithm, which is commonly used to solve heterogeneous task scheduling problem. Based on our extensive simulation study, we observe that our algorithms exhibit excellent performance and demonstrate that considering data allocation in task scheduling is significant for saving energy. We conduct experiments on two heterogeneous multiprocessor systems.Given a heterogeneous hybrid memory multiprocessor system that consists of n heterogeneous processors P 1 ; P 2 ; : : : ; P n , and each processor P i contains a hybrid memory, which is equipped with DRAM i and SSD i . The access time and access energy of each processor in accessing a unit data In DAA_TS algorithm, we first use the HEFT algorithm to initialize the schedule. And then, we propose Data-Aware Algorithm, which is denoted in Algorithm 3 to find the share data and the single data. The share data and the single data are denoted as share_d and si ngle_d , respectively. We can save much more cost to first assign share_d , because the memory store share_d will be accessed more than once.After identifying the type of data, we propose Data Assignment Algorithm described in Algorithm 4. In Algorithm 4, we build two queues, one is a memory queue denoted as M and other one is a Then, each single data is always placed in the processor to access the fastest and spatial memory. As a result, we draw the tasks and data schedule scheme. However, this result was not the most ideal. In order to satisfy the time constraint and the minimal energy consumption with the resource constraint, the DAA_TS algorithm is also contained in the following optimization algorithm to reduce the energy consumption EC and the completion time T .At first, we change tasks and data position to achieve the energy consumption of minimal mi n_E. And then, the task scheduling and data assignment problem is solved, if the total completion time T satisfies the time constraint S . Otherwise, for reaching the time constraint S requirement, tasks and data should be reallocated. After T < S, we will adopt tasks migration. While reallocating a set of data D, until the energy consumption cannot be reduced any more. Otherwise, we should reallocate some tasks and some data to guarantee the finish time...

show abstract

“…To the best of our knowledge, most studies in the literature are based on the exact values of the actual conditions. In a dynamic environment, especially for CPD process, it is difficult to guarantee the precise value of the task execution time, the completion time, and the delivery 5 . On the other hand, task allocating is flexible and learnable, that is, the execution time is variable and controllable.…”

Section: Introductionmentioning

confidence: 99%

A Robust Predictive–Reactive Allocating Approach, Considering Random Design Change in Complex Product Design Processes

Su¹,

Meng²,

Liu³

2018

IJCIS

View full text Add to dashboard Cite

In the highly dynamic complex product design process, task allocations recovered by reactive allocating decisions are usually subject to design changes. In this paper, a robust predictive-reactive allocating approach considering possible disruption times is proposed, so that it can absorb the disruption in the executing process and utilize the limited capacity of resource more effectively. Four indexes (Makespan, stability, robustness, and compression cost) are used to measure the quality of the proposed method. To illustrate the novel allocating idea, we first assign tasks to resources with the objective of a trade-off between the overall execution time and the overall design cost, which can transform the problem into a non-identical parallel environment. Then, the probability distribution sequencing (PDS) method combining with inserting idle time (IIT) is proposed to generate an original-predictive allocation. A match-up time strategy is considered to match up with the initial allocation at some point in the future. The relationship between the minimum match-up time and the compression cost is analyzed to find the optimal matchup time. Our computational results show that the proposed sequencing method is better than the shortest processing time (SPT) which is a common sequencing way mentioned in the literature. The robust predictive-reactive allocating approach is sensitive to the design change, which is helpful to reduce the reallocating cost and keep the robustness and stability.

show abstract

Energy-Aware Data Allocation and Task Scheduling on Heterogeneous Multiprocessor Systems With Time Constraints

Cited by 63 publications

References 33 publications

Efficient Energy Consumption Scheduling: Towards Effective Load Leveling

Efficient Energy Consumption Scheduling: Towards Effective Load Leveling

Data‐aware task scheduling on heterogeneous hybrid memory multiprocessor systems

A Robust Predictive–Reactive Allocating Approach, Considering Random Design Change in Complex Product Design Processes

Contact Info

Product

Resources

About