k-Server Optimal Task Scheduling Problem with Convex Cost Function
WIOPT '05 Proceedings of the Third International Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks
On bounding energy consumption in dynamic, embedded real-time systems
Proceedings of the 2006 ACM symposium on Applied computing
Utility Accrual Real-Time Scheduling Under the Unimodal Arbitrary Arrival Model with Energy Bounds
IEEE Transactions on Computers
Stochastic DVS-based dynamic power management for soft real-time systems
Microprocessors & Microsystems
Optimal service level allocation in environmentally powered embedded systems
Proceedings of the 2009 ACM symposium on Applied Computing
Utility-based scheduling for grid computing under constraints of energy budget and deadline
Computer Standards & Interfaces
An energy management framework for energy harvesting embedded systems
ACM Journal on Emerging Technologies in Computing Systems (JETC)
Dynamic power management in environmentally powered systems
Proceedings of the 2010 Asia and South Pacific Design Automation Conference
PATMOS'05 Proceedings of the 15th international conference on Integrated Circuit and System Design: power and Timing Modeling, Optimization and Simulation
An experimental evaluation of real-time DVFS scheduling algorithms
Proceedings of the 5th Annual International Systems and Storage Conference
Hi-index | 0.01 |
In this paper, we explore the feasibility and performanceoptimization problems for real-time systems that must remainfunctional during an operation/mission with a fixed, initialenergy budget. We show that the feasibility problem is NP-Hardin the context of systems with Dynamic Voltage Scaling(DVS) capability and discrete speed levels. Then, we focuson energy-constrained periodic task systems where theavailable energy budget is not sufficient to meet all thedeadlines. We propose techniques to maximize the totalnumber of deadlines met and the total reward (utility) whileguaranteeing the completion of the mission and a minimumperformance for each task. We consider separately: (i)systems with or without DVS capability, and (ii) off-line(static) and on-line (dynamic) solutions to select mostvaluable jobs for execution. We also discuss the tractabilityof the involved optimization problems. Our on-linealgorithms combine job promotion, job demotion and speedreduction techniques to maximize the system performancewhile guaranteeing the completion of the mission. Weevaluate our schemes through simulations and show that theon-line schemes can yield significant performanceimprovements over static solutions.