東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Stochastic modeling of a power-manag...

Qui, Qinru.

Linked to FindBook

Google Book

Amazon

博客來

Stochastic modeling of a power-managed system: Construction and optimization.

Record Type:	Language materials, printed : Monograph/item
Title/Author:	Stochastic modeling of a power-managed system: Construction and optimization./
Author:	Qui, Qinru.
Description:	173 p.
Notes:	Adviser: Massoud Pedram.
Contained By:	Dissertation Abstracts International62-09B.
Subject:	Engineering, Electronics and Electrical. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3027767
ISBN:	0493398244

Stochastic modeling of a power-managed system: Construction and optimization.
Qui, Qinru.

Stochastic modeling of a power-managed system: Construction and optimization. - 173 p.

Adviser: Massoud Pedram.

Thesis (Ph.D.)--University of Southern California, 2001.

Low power has become an increasingly important design target due to the popularity of battery-operated portable devices. Low power consumption is also important in the design of high performance circuits because it results in a decrease in the packaging and cooling cost and an increase in the circuit reliability. Power optimization techniques have been studied at all levels and layout levels. This thesis focuses on system-level power optimization based on dynamic power management.

ISBN: 0493398244Subjects--Topical Terms:

626636
Engineering, Electronics and Electrical.

Stochastic modeling of a power-managed system: Construction and optimization.
LDR:02960nam 2200301 a 45 001 932464
005 20110505
008 110505s2001 eng d
020 $a 0493398244
035 $a (UnM)AAI3027767
035 $a AAI3027767
040 $a UnM $c UnM
100 1 $a Qui, Qinru. $3 1256206
245 1 0 $a Stochastic modeling of a power-managed system: Construction and optimization.
300 $a 173 p.
500 $a Adviser: Massoud Pedram.
500 $a Source: Dissertation Abstracts International, Volume: 62-09, Section: B, page: 4149.
502 $a Thesis (Ph.D.)--University of Southern California, 2001.
520 $a Low power has become an increasingly important design target due to the popularity of battery-operated portable devices. Low power consumption is also important in the design of high performance circuits because it results in a decrease in the packaging and cooling cost and an increase in the circuit reliability. Power optimization techniques have been studied at all levels and layout levels. This thesis focuses on system-level power optimization based on dynamic power management.
520 $a The goal of dynamic power management is to reduce the power consumption of an electronic system by putting system components into different states, each representing certain performance and power consumption levels. The power management policy determines the type and timing of these transitions based on the system history, workload, and performance constraints.
520 $a system. Two modeling techniques are proposed. The first approach is targeted at a simple power-managed system, which contains only one service provider, a priority request queue, and an outside request generator. The power management of the system is modeled and solved as an optimization problem based on the theory of continuous-time Markov decision processes. The second approach is targeted at a complex power-managed system, which contains multiple service providers and complex interactions among them. The power management of the system is modeled using the controllable generalized stochastic Petri Net. The optimal policy is solved exactly using linear programming or near-exactly using the “policy iteration” algorithm. Optimization algorithms for both deterministic control policy and randomized control policy are presented.
520 $a Finally, the stochastic power management policy is compared with existing heuristic methods for different workload statistics. Experimental results show that the power management methods based on stochastic theories outperform heuristic methods in terms of higher saving in power dissipation while achieving a constraint on the system performance.
590 $a School code: 0208.
650 4 $a Engineering, Electronics and Electrical. $3 626636
690 $a 0544
710 2 0 $a University of Southern California. $3 700129
773 0 $t Dissertation Abstracts International $g 62-09B.
790 $a 0208
790 1 0 $a Pedram, Massoud, $e advisor
791 $a Ph.D.
792 $a 2001
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3027767