Page 1 |
Save page Remove page | Previous | 1 of 275 | Next |
|
small (250x250 max)
medium (500x500 max)
Large (1000x1000 max)
Extra Large
large ( > 500x500)
Full Resolution
All (PDF)
|
This page
All
|
DETERMINISTIC MATHEMATICAL OPTIMIZATION IN STOCHASTIC NETWORK CONTROL by Longbo Huang A Dissertation Presented to the FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Ful llment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (ELECTRICAL ENGINEERING) August 2011 Copyright 2011 Longbo Huang
Object Description
Title | Deterministic mathematical optimization in stochastic network control |
Author | Huang, Longbo |
Author email | longbo.huang@gmail.com;longbohu@usc.edu |
Degree | Doctor of Philosophy |
Document type | Dissertation |
Degree program | Electrical Engineering |
School | Viterbi School of Engineering |
Date defended/completed | 2011-05-06 |
Date submitted | 2011-07-06 |
Date approved | 2011-07-06 |
Restricted until | 2011-07-06 |
Date published | 2011-07-06 |
Advisor (committee chair) | Neely, Michael J. |
Advisor (committee member) |
Caire, Giuseppe Jain, Rahul Kempe, David Krishnamachari, Bhaskar Wierman, Adam |
Abstract | In this thesis, we extend the recently developed Lyapunov optimization technique (also known as Max-Weight or Backpressure) for stochastic queueing networks in two important directions: (1) guaranteeing small network delay; and (2) resolving underflows. ❧ To achieve our objective, we first establish an explicit connection between the Lyapunov technique and a randomized dual subgradient method. Based on this connection, we develop a novel exponential attraction result, which states that the network queue backlog under a Lyapunov algorithm deviates from a certain fixed point with a probability that decreases exponentially in the deviation distance. Inspired by the exponential attraction result, we develop three delay-efficient algorithms and show that they achieve near-optimal utility-delay tradeoffs for a general class of multi-hop communication networks. One of the algorithms has also been implemented on a sensor network testbed and was shown to be able to guarantee very small network delay in practical systems. ❧ We later consider the problem of resolving underflows in general complex network scheduling problems. In this case, we propose the weight perturbation technique and develop the Perturbed Max-Weight algorithm (PMW). We show that PMW effectively resolves underflow constraints without sacrificing utility performance. We then apply the perturbation technique to construct utility optimal scheduling algorithms for two important classes of networks -- stochastic processing networks and energy harvesting networks. ❧ The results developed in this thesis highlight the importance of Lagrange multiplier engineering in queueing networks. Specifically, our results show that the queues under the Lyapunov technique indeed correspond to the Lagrange multiplier values under the randomized dual subgradient method. This not only helps us better understand the Lyapunov technique, but also gives us general guidelines on how should one design its algorithm to achieve the desire properties of the queues. |
Keyword | stochastic network control; network delay; queueing; Lyapunov analysis; processing networks; energy harvesting networks |
Language | English |
Part of collection | University of Southern California dissertations and theses |
Publisher (of the original version) | University of Southern California |
Place of publication (of the original version) | Los Angeles, California |
Publisher (of the digital version) | University of Southern California. Libraries |
Provenance | Electronically uploaded by the author |
Type | texts |
Legacy record ID | usctheses-m |
Contributing entity | University of Southern California |
Rights | Huang, Longbo |
Physical access | The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given. |
Repository name | University of Southern California Digital Library |
Repository address | USC Digital Library, University of Southern California, University Park Campus MC 7002, 106 University Village, Los Angeles, California 90089-7002, USA |
Repository email | cisadmin@lib.usc.edu |
Archival file | uscthesesreloadpub_Volume71/etd-HuangLongb-49.pdf |
Description
Title | Page 1 |
Contributing entity | University of Southern California |
Repository email | cisadmin@lib.usc.edu |
Full text | DETERMINISTIC MATHEMATICAL OPTIMIZATION IN STOCHASTIC NETWORK CONTROL by Longbo Huang A Dissertation Presented to the FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Ful llment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (ELECTRICAL ENGINEERING) August 2011 Copyright 2011 Longbo Huang |