國立高雄大學圖資館 |

Language: English

Back

Jointly optimal network performance :A cross-layer approach

Record Type:	Electronic resources : Monograph/item
Title/Author:	Jointly optimal network performance :
Reminder of title:	A cross-layer approach
Author:	O'Neill, Daniel C.
Description:	125 p.
Notes:	Advisers: Stephen Boyd; Nick Bambos.
Notes:	Source: Dissertation Abstracts International, Volume: 65-04, Section: B, page: 2018.
Contained By:	Dissertation Abstracts International65-04B.
Subject:	Engineering, Electronics and Electrical.
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3128449
ISBN:	0496757113

Jointly optimal network performance :A cross-layer approach
O'Neill, Daniel C.

Jointly optimal network performance :A cross-layer approach [electronic resource] - 125 p.

Advisers: Stephen Boyd; Nick Bambos.

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

The first of three network management algorithms is the Approximate Projected Sub-Gradient method. This method re-expresses the formal optimization problem in eigenvalue form and presents a method of solution based on sub-gradients to the network performance metric. The eigenvalue is interpreted as the utilization of the network and an associated left eigenvector interpreted as capturing the effects of link interference. A theory based heuristic is developed to more easily implement the approach.

ISBN: 0496757113Subjects--Topical Terms:

226981
Engineering, Electronics and Electrical.

Jointly optimal network performance :A cross-layer approach
LDR:03265nmm _2200337 _450 001 162609
005 20051017073504.5
008 230606s2004 eng d
020 $a 0496757113
035 $a 00149110
035 $a 162609
040 $a UnM $c UnM
100 0 $a O'Neill, Daniel C. $3 227753
245 1 0 $a Jointly optimal network performance : $b A cross-layer approach $h [electronic resource]
300 $a 125 p.
500 $a Advisers: Stephen Boyd; Nick Bambos.
500 $a Source: Dissertation Abstracts International, Volume: 65-04, Section: B, page: 2018.
502 $a Thesis (Ph.D.)--Stanford University, 2004.
520 # $a The first of three network management algorithms is the Approximate Projected Sub-Gradient method. This method re-expresses the formal optimization problem in eigenvalue form and presents a method of solution based on sub-gradients to the network performance metric. The eigenvalue is interpreted as the utilization of the network and an associated left eigenvector interpreted as capturing the effects of link interference. A theory based heuristic is developed to more easily implement the approach.
520 # $a The practical problem of managing switch power consumption in wireless networks is also considered. A method to optimally manage the switch, termed Minimum Cost Matching/Adaptive Framing, is presented. The method is asymptotically stable and can be adapted to a wide array of constraints.
520 # $a The second method is the Congestion Bandwidth Product method. This method is based on the dual to the original optimization problem. The result is a congestion minimizing algorithm based on a sub-gradient to the dual function. A key point is that the queues in the network behave as if they were a sub-gradient minimization method.
520 # $a The thesis develops an analytical framework for modelling a wireless network capable of supporting a variety of different media types and defines a variety of network performance measures. Interference effects are modelled, as are expressions for link data rates. A key point is the convexity of the "rate region" for both the forward and the reverse networks. The behavior of sources are modelled by system performance metrics or by utility functions. The network control problem is formulated as an optimization problem.
520 # $a The third method is based on Geometric Programming, a centralized convex optimization technique. The approach is efficient and well suited to networks with complex QoS constraints.
520 # $a This thesis focuses on the problem of managing or of controlling a wireless network. Three different approaches are presented. These approaches operate at both the physical and the transport layer and so are termed cross-layer network control techniques. In addition, the practical problem of managing a packet switch in such a network is considered.
590 $a School code: 0212.
650 # 0 $a Engineering, Electronics and Electrical. $3 226981
690 $a 0544
710 0 # $a Stanford University. $3 212607
773 0 # $g 65-04B. $t Dissertation Abstracts International
790 $a 0212
790 1 0 $a Bambos, Nick, $e advisor
790 1 0 $a Boyd, Stephen, $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=3128449 $z http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3128449