國立高雄大學圖資館 |

Language: English

Back

Decomposition and fast distributed iterations for optimization of networked systems

Record Type:	Electronic resources : Monograph/item
Title/Author:	Decomposition and fast distributed iterations for optimization of networked systems
Author:	Xiao, Lin.
Description:	147 p.
Notes:	Adviser: Stephen P. Boyd.
Notes:	Source: Dissertation Abstracts International, Volume: 65-04, Section: B, page: 2027.
Contained By:	Dissertation Abstracts International65-04B.
Subject:	Engineering, Electronics and Electrical.
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3128500
ISBN:	0496757628

Decomposition and fast distributed iterations for optimization of networked systems
Xiao, Lin.

Decomposition and fast distributed iterations for optimization of networked systems [electronic resource] - 147 p.

Adviser: Stephen P. Boyd.

Thesis (Ph.D.)--Stanford University, 2004.

It is shown that some formulations of the weights-for-fastest-convergence problem are tractable, and can be formulated as semidefinite programs. Special structure of the problems are exploited to gain efficiency in interior-point methods, and to develop specialized subgradient methods to handle problems with far larger scales. The performance of algorithms with optimally tuned weights can be significantly better than with weights given by simple heuristics.

ISBN: 0496757628Subjects--Topical Terms:

226981
Engineering, Electronics and Electrical.

Decomposition and fast distributed iterations for optimization of networked systems
LDR:02942nmm _2200325 _450 001 162622
005 20051017073506.5
008 230606s2004 eng d
020 $a 0496757628
035 $a 00149123
035 $a 162622
040 $a UnM $c UnM
100 0 $a Xiao, Lin. $3 227766
245 1 0 $a Decomposition and fast distributed iterations for optimization of networked systems $h [electronic resource]
300 $a 147 p.
500 $a Adviser: Stephen P. Boyd.
500 $a Source: Dissertation Abstracts International, Volume: 65-04, Section: B, page: 2027.
502 $a Thesis (Ph.D.)--Stanford University, 2004.
520 # $a It is shown that some formulations of the weights-for-fastest-convergence problem are tractable, and can be formulated as semidefinite programs. Special structure of the problems are exploited to gain efficiency in interior-point methods, and to develop specialized subgradient methods to handle problems with far larger scales. The performance of algorithms with optimally tuned weights can be significantly better than with weights given by simple heuristics.
520 # $a The first topic concerns a cross-layer optimization problem in wireless communication networks---simultaneous routing and resource allocation. This dissertation introduces a convex optimization formulation of this problem, and exploit the layered architecture of wireless networks via dual de composition method, to develop a distributed pricing algorithm on the capacities of communication links to optimally coordinate routing and resource allocation.
520 # $a The second topic concerns designing fast distributed iterations via semidefinite programming. A class of distributed iterative algorithms on a graph are considered, where the local computation or communication at each step involve some weights or parameters on the edges that determine the speed of convergence of the algorithms. Three specific problems are studied under a common computational framework: finding the fastest mixing Markov chain on a graph, designing fast linear iterations for distributed average consensus, and designing fast distributed algorithms for an optimal resource allocation problem.
520 # $a This dissertation studies several optimization problems arising in networked systems (with an underlying graph structure), focusing on their convex optimization formulations and exploitation of problem structure to develop efficient, specialized solution methods and distributed algorithms. There are two major topics covered in this thesis.
590 $a School code: 0212.
650 # 0 $a Engineering, Electronics and Electrical. $3 226981
650 # 0 $a Computer Science. $3 212513
650 # 0 $a Engineering, System Science. $3 227118
690 $a 0544
690 $a 0790
690 $a 0984
710 0 # $a Stanford University. $3 212607
773 0 # $g 65-04B. $t Dissertation Abstracts International
790 $a 0212
790 1 0 $a Boyd, Stephen P., $e advisor
791 $a Ph.D.
792 $a 2004
856 4 0 $u http://libsw.nuk.edu.tw/login?url=http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3128500 $z http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3128500