Martin, William J
Semidefinite programming is a recently developed branch of convex optimization which optimizes a linear function subject to nonlinear constraints, the most important of which require that a combination of symmetric matrices be positive semidefinite. Semidefinite programming has a broad applicability and algorithmic efficiency, making it very appealing for use in all kinds of areas of study. This project begins by focusing on the understanding of semidefinite programming and its methods for numerical solution. It then surveys the vast and exciting applications of semidefinite programming and investigates the Sensor Network Localization Problem (SNLP). Lastly, it provides a tutorial for implementing the use of online tools and the computer codes I developed to solve the SNLP.
Worcester Polytechnic Institute
Major Qualifying Project
All authors have granted to WPI a nonexclusive royalty-free license to distribute copies of the work, subject to other agreements. Copyright is held by the author or authors, with all rights reserved, unless otherwise noted.