The views expressed in this thesis are those of the author and do not reflect the official policy or position of the Department of Defense or the U.S. Government. / Determining the best base stationing for military units can be modeled as a capacitated facility location problem with sole sourcing and multiple resource categories. Computational experience suggests that cross decomposition, a unification of Benders Decomposition and Lagrangean relaxation, is superior to other contemporary methods for solving capacitated facility location problems. Recent research extends cross decomposition to pure integer prograrnming problems with explicit application to capacitated facility location problems with sole sourcing; however, this research offers no computational experience. This thesis implements two cross decomposition algorithms for the capacitated facility location problem with sole sourcing and compares these decomposition algorithms with branch and bound methods. For some problems tested, cross decomposition obtains better solutions in less time; however, cross decomposition does not always perform better man branch and bound due to the time required to obtain the cross decomposition bound that is theoretically superior to other decomposition bounds.

No description available.

Mobile phones are becoming a primary platform for information access. More and more people use their mobile devices as one of their major communication access tools. Commuters are increasingly carrying their mobile devices with them almost everywhere. Mobile devices fit perfectly into an ideal environment for realizing ubiquitous computing. A major aspect of ubiquitous computing is context-aware applications where the applications collect information about the environment that the user is in and use this information to achieve their goals or improve performance. The location of the device is one of the most important pieces of context information. Location awareness makes certain applications possible, e.g., recommending nearby businesses and tracking estimated routes, and greatly improves the performance of other applications, for example it can be associated with automobile navigation devices. A feature available to mobile applications in the Android platform makes it possible to determine a device's location without any additional hardware or sensor mechanisms, by simply using the native capability of the built-in wireless network card. Since the release of Android systems, there have been numerous applications developed to introduce new ways of tracking locations. Recently, there have been many papers on location estimation leveraging ubiquitous wireless networks. In this thesis, we develop an Android application to collect useful Wi-Fi information without registering a location listener with a network-based provider, such as Wi-Fi connections or data connections. Therefore it provides a passive, privacy-preserving, non-intrusive and power-saving way of achieving location awareness to Android mobile users. Accurate estimation of the location information can bring a more contextual experience to mobile users. We save the passively collected data of the IDs of Wi-Fi access points and the received signal strengths to a database in order to help us structure the data and analyse it. We employ some heuristics to infer the location information from the data. Our work presents a location tracking technique mainly based on Basic Service Set identification (BSSID) and/or Received Signal Strength Indicator (RSSI) using Wi-Fi information. It falls into one of the most active fields in mobile application development --location-based or location-aware applications.

We propose a novel network-based approach for location estimation in social media that integrates evidence of the social tie strength between users for improved location estimation. Concretely, we propose a location estimator – FriendlyLocation– that leverages the relationship between the strength of the tie between a pair of users, and the distance between the pair. Based on an examination of over 100 million geo-encoded tweets and 73 million Twitter user profiles, we identify several factors such as the number of followers and how the users interact that can strongly reveal the distance between a pair of users. We use these factors to train a decision tree to distinguish between pairs of users who are likely to live nearby and pairs of users who are likely to live in different areas. We use the results of this decision tree as the input to a maximum likelihood estimator to predict a user’s location. We find that this proposed method significantly improves the results of location estimation relative to a state-of-the-art technique. Our system reduces the average error distance for 80% of Twitter users from 40 miles to 21 miles using only information from the user’s friends and friends-of-friends, which has great significance for augmenting traditional social media and enriching location-based services with more refined and accurate location estimates.

The capacitated facility location problem is a well known problem in combinatorial optimization and operations research. In it, we are given a set of clients and a set of possible facility locations. Each client has a certain demand that needs to be satisfied from open facilities, without exceeding their capacity. Whenever we open a facility we incur in a corresponding opening cost. Whenever demand is served, we incur in an assignment cost; depending on the distance the demand travels. The goal is to open a set of facilities that satisfy all demands while minimizing the total opening and assignment costs. In this thesis, we present two novel LP-based approximation algorithms for the capacitated facility location problem. The first algorithm is based on LP-rounding techniques, and is designed for the special case of the capacitated facility location problem where capacities are uniform and assignment costs are given by a tree metric. The second algorithm follows a primal-dual approach, and works for the general case. For both algorithms, we obtain an approximation guarantee that is linear on the size of the problem. To the best of our knowledge, there are no LP-based algorithms known, for the type of instances that we focus on, that achieve a better performance.

Thesis (MEng) -- University of South Australia, 1994

Thesis (PhD)--University of South Australia, 1999

A research area that has recently gained great interest is the development of network architectures relating to the tracking of wireless VoIP devices. This is particularly so for architectures based on the popular Session Initiation Protocol (SIP). Previous work, however, in this area does not consider the impact of combined VoIP and tracking on the capacity and call set-up time of the architectures. Previous work also assumes that location information is always available from sources such as GPS, a scenario that rarely is found in practice. The inclusion of multiple positioning systems in tracking architectures has not been hitherto explored. It is the purpose of this thesis to design and test SIP-based architectures that address these key issues. Our first main contribution is the development of a tracking-only SIP based architecture. This architecture is designed for intermittent GPS availability, with wireless network tracking as the back-up positioning technology. Such a combined tracking system is more conducive with deployment in real-world environments. Our second main contribution is the development of SIP based tracking architectures that are specifically aimed at mobile wireless VoIP systems. A key aspect we investigate is the quantification of the capacity constraints imposed on VoIP-tracking architectures. We identify such capacity limits in terms of SIP call setup time and VoIP QoS metrics, and determine these limits through experimental measurement and theoretical analyses. Our third main contribution is the development of a novel SIP based location tracking architecture in which the VoIP application is modified. The key aspect of this architecture is the factor of two increase in capacity that it can accommodate relative to architectures utilizing standard VoIP. An important aspect of all our tracking architectures is the Tracking Server. This server supplies the location information in the event of GPS unavailability. A final contribution of this thesis is the development of novel particle-filter based tracking algorithms that specifically address the GPS intermittency issue. We show how these filters interact with other features of our SIP based architectures in a seamless fashion.

Thesis (Ph.D.)--McMaster University, 2004. / Supervisor: George O. Wesolowsky. Includes bibliographical references (leaves 120-125). Also available online.

Thesis (Ph.D.)--McMaster University, 2005. / Includes bibliographical references. Also available online.