A Queuing Theoretic Approach to Processor Power Adaptation for Video Decoding Systems

Foo, Brian; Schaar, Mihaela van der

doi:10.1109/tsp.2007.906737

Cited by 6 publications

(25 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In real time video transmission, this can be accomplished by having the encoder send complexity specifications, such as the mean and variance of each job class for each video scene, prior to transmitting the corresponding frames [8].…”

Section: Consideration Of Stochastic Complexitymentioning

confidence: 99%

“…In our experiment, we revised power models of laEDF [3] and queuing based algorithms [8] to consider leakage power. Also we revised these algorithms to consider sleep mode for a fair comparison.…”

Section: Optimality Studymentioning

confidence: 99%

“…A wide variety of DVS algorithms have been proposed for delay-sensitive applications [3][4][5][6][7][8]. Some DVS algorithms perform optimization myopically over only one or two tasks, such that the processor power level is determined on the fly to meet imminent (soft) deadlines while considering either the worst case execution time (WCET), or the average case execution time (ACET) [4] [6].…”

Section: Introductionmentioning

confidence: 99%

“…Based on the workload traces collected during execution time, we solve the offline LP problem to obtain the lower energy bound for DVS algorithms. A thorough investigation of video decoding results (where many video sequences are decoded at many different bit rates) shows that, under the same zero miss rate, laEDF [3] consumes approximately 15% more energy than the optimal solution, and the queuing based algorithm in [8] consumes approximately 4% more than the optimal solution.…”

Section: Introductionmentioning

confidence: 99%

“…Experimental results from real-time video decoding (where workloads are highly time-varying) indicate that SLP/r consumes less than 1% more energy than the optimal DVS solution while dropping only 0.1% of all scheduled decoding jobs, while the best existing algorithm (the queuing-based algorithm 2 in [8]) consumes roughly 3% more energy than SLP/r at the same miss rate. Importantly, SLP/r has only a small increase in complexity, and its relative complexity scales down when supporting increasingly computationally intensive future multimedia applications.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Optimality and improvement of dynamic voltage scaling algorithms for multimedia applications

Cao

Foo

et al. 2008

Proceedings of the 45th Annual Design Automation Conference

Self Cite

View full text Add to dashboard Cite

*The time-varying workload for multimedia applications poses a great challenge for the efficient performance of dynamic voltage scaling (DVS) algorithms. While many DVS algorithms have been proposed for real-time applications, there does not yet exist a systematic method for evaluating the optimality of such DVS algorithms. In this paper, we propose an offline linear programming (LP) method to determine the minimum energy consumption for processing multimedia tasks under stringent delay deadlines. Based on this lower bound, we evaluate the efficiency of various existing DVS algorithms. Furthermore, we modify the LP formulation to construct an online robust sequential linear programming DVS algorithm for real-time multimedia processing. Simulation results from decoding over a wide range of video sequences shows that on average, our online algorithm consumes less than 1% more energy than the optimal lower bound while dropping only 0.1% of all scheduled decoding jobs, while the existing best algorithm consumes roughly 3% more energy at the same miss rate.

show abstract

Section: Consideration Of Stochastic Complexitymentioning

confidence: 99%

Section: Optimality Studymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Optimality and improvement of dynamic voltage scaling algorithms for multimedia applications

Cao

Foo

et al. 2008

Proceedings of the 45th Annual Design Automation Conference

Self Cite

View full text Add to dashboard Cite

show abstract

Energy-Aware Scheduling for Sensor Node Platforms

Tak

Kim

et al. 2014

2014 15th International Conference on Parallel and Distributed Computing, Applications and Technologies

View full text Add to dashboard Cite

This paper primarily presents a technique to integrate energy-aware real-time scheduling into batterypowered sensor node platforms. The proposed scheduling technique attempts to achieve energy savings while meeting timing requirements of real-time tasks as well as improving the response time of non-real-time tasks. The proposed energy-aware real-time scheduling technique is addressed from two perspectives: a dynamic voltage scaling (DVS)-based scheduling approach and a combined real-time and non-real-time scheduling approach. From the perspective of DVS-based scheduling approach, it reduces the processor frequency and still ensures that no tasks miss their timing constraints. From the perspective of a combined real-time and non-real-time scheduling scheme, tasks are allowed to have timing constraints. Consequently, realtime tasks need to be processed in a timely fashion and delivered to meet their timing constraints, and non-real-time tasks need to be quickly processed to deliver a fast response. An experimental evaluation shows that the proposed scheduling technique yields efficient performance in terms of low energy consumption and a fast average response time for non-real-time tasks, whilst meeting the timing constraints of real-time tasks.

show abstract

Optimality and Improvement of Dynamic Voltage Scaling Algorithms for Multimedia Applications

Cao

Foo

et al. 2010

IEEE Trans. Circuits Syst. I

Self Cite

View full text Add to dashboard Cite

The time-varying workload for multimedia applications poses a great challenge for the efficient performance of dynamic voltage scaling (DVS) algorithms. While many DVS algorithms have been proposed for real-time applications, there does not yet exist a systematic method for evaluating the optimality of such DVS algorithms. In this paper, we propose an offline linear programming (LP) method to determine the minimum energy consumption for processing multimedia tasks under stringent delay deadlines. Based on this lower bound, we evaluate the efficiency of various existing DVS algorithms. Furthermore, we modify the LP formulation to construct an online robust sequential linear programming DVS algorithm for real-time multimedia processing. Simulation results from decoding over a wide range of video sequences shows that on average, our online algorithm consumes less than 1% more energy than the optimal lower bound while dropping only 0.1% of all scheduled decoding jobs, while the existing best algorithm consumes roughly 3% more energy at the same miss rate.

show abstract

A Queuing Theoretic Approach to Processor Power Adaptation for Video Decoding Systems

Cited by 6 publications

References 36 publications

Optimality and improvement of dynamic voltage scaling algorithms for multimedia applications

Optimality and improvement of dynamic voltage scaling algorithms for multimedia applications

Energy-Aware Scheduling for Sensor Node Platforms

Optimality and Improvement of Dynamic Voltage Scaling Algorithms for Multimedia Applications

Contact Info

Product

Resources

About