This thesis investigates the automatic identification of the location of doc- uments. This process of geolocation aids in toponym resolution, document summarization, and geographic-based marketing. I focus on minimally su- pervised methods to examine both the lexical similarities and the geographic similarities between documents. This method predicts the location of a doc- ument as a single point on the earth’s surface. Three data sets are used to evaluate this method: a set of geotagged Wikipedia articles and two sets of Twitter feeds. For Wikipedia, the combined method obtains a median error of 12.1 kilometers and an improvement in mean error to 164 kilometers. The large Twitter data shows the greatest improvement from this method with a median error of 333 kilometers, down from the previous best of 463 kilometers. / text

Approved for public release / The proliferation of Fourth Generation (4G) mobile wireless networks has led to an increased demand for position-based services. In many densely populated environments, there is an abundance of base stations by which a mobile station may be passively geographically located (geolocated) using various techniques that require multiple base stations. Areas in which base station density is sparse may not possess the requisite number of base stations to perform such techniques and, therefore, require different methods by which to geolocate mobile stations. In this thesis, we present a passive geolocation scheme that only requires observation of the initial ranging information exchange between a mobile station and a single base station in order to determine a position estimate for the mobile station. The scheme is specifically applied to the Institute of Electrical and Electronics Engineers (IEEE) 802.16 Worldwide Interoperability for Microwave Access (WiMAX) 4G standard. The method is validated through computer simulation and field experimentation in an Alcatel-Lucent IEEE 802.16e-2005 deployed network. The mean geolocation error resulting from simulation and experimentation was 85 meters, which is a degradation of 58 meters compared to a three-base station scenario. / Captain, United States Marine Corps

Recently, indoor geolocation technologies has been attracting tremendous attention. For indoor environments, the fine time resolution of ultra-wideband (UWB) signals enables the potential of accurate distance measurement of the direct path (DP) between a number of reference sources and the people or assets of interest. However, Once the DP is not available or is shadowed, substantial errors will be introduced into the ranging measurements, leading to large localization errors when measurements are combined from multiple sources. The measurement accuracy in undetected direct path (UDP) conditions can be improved in some cases by exploiting the geolocation information contained in the indirect path measurements. Therefore, the dynamic spatial behavior of paths is an important issue for positioning techniques based on TOA of indirect paths. The objectives of this thesis are twofold. The first is to analyze the sensitivity of TOA estimation techniques based on TOA of the direct path. we studied the effect of distance, bandwidth and multipath environment on the accuracy of various TOA estimation techniques. The second is to study the sensitivity of multipath parameters pertinent to TOA estimation techniques based on the TOA of the indirect paths. We mainly looked into the effect of distance, bandwidth, threshold for picking paths, and multipath environment on the number of multipath components(MPCs) and path persistency. Our results are based on data from a new measurement campaign conducted on the 3rd floor of AK laboratory. For the TOA estimation techniques based on DP, the line of sight (LOS) scenario provides greatest accuracy and these TOA estimation techniques are most sensitive to bandwidth availability in obstructed line of sight (OLOS) scenario. All the TOA estimation algorithms perform poorly in the UDP scenario although the use of higher bandwidth can reduce the ranging error to some extent. Based on our processed results, The proposal for selecting the appropriate TOA estimation technique with certain constrains is given. The sensitivity study of multipath parameters pertinent to indirect-path-based TOA estimation techniques shows that the number of MPCs is very sensitive to the threshold for picking paths and to the noise threshold. It generally decreases as the distance increase while larger bandwidth always resolves more MPCs. The multipath components behave more persistently in line of sight (LOS) and obstructed line of sight (OLOS) scenarios than in UDP scenarios, and the use of larger bandwidth and higher threshold for picking paths also result in more persistent paths.

The Federal Communications Commission (FCC) has adopted regulations requiring wireless communication service providers to provide position location (PL) information for a user requesting E-911 service. The Time Difference of Arrival (TDOA) technique is one of the most promising position location techniques for cellular-type wireless communication systems. The IS-95 Code Division Multiple Access (CDMA) system is a popular choice for the companies deploying new cellular and PCS systems in North America. Hence, the feasibility of TDOA techniques in CDMA systems is an important issue for position location in the wireless systems of the future. This thesis analyzes the performance of TDOA techniques in the CDMA systems. A comparison and assessment of different algorithms for finding the time difference estimates and for solving the hyperbolic equations generated by those estimates has been made. This research also considers a measure of accuracy for TDOA position location method which is shown to be more suitable for CDMA systems and more closely matches to the FCC requirements. Among the other contributions is a proposed method to perform cross-correlations to identify only the desired user's TDOA in a multiuser environment. This thesis also evaluates the feasibility and accuracy of TDOA techniques under varying system conditions that might be encountered in real situations. This includes varying conditions of Additive White Gaussian Noise (AWGN), Multiple Access Interference (MAI), power control and loading. The effect of the mobile position and of different arrangement of base stations on TDOA accuracy is also studied. Performance comparison in AWGN and Rayleigh fading channels is made. The feasibility of using increased power levels for the 911 user in combination with interference cancellation is also studied. The effect of using a single stage of parallel interference cancellation at neighboring cell sites has also been explored. Non-ideal situations such as imperfect power control in CDMA operation has also been investigated in the context of position location. This thesis also suggests amethod to correct TDOA estimation errors in CDMA. It is shown that this improvement can give greatly improved performance even under worst-case situations. Performance comparison of results with and without that modification has also been made under various conditions. / Master of Science

Dissertation (Ph.D.)--Worcester Polytechnic Institute. / Keywords: Time-of-arrival based ranging, indoor geolocation, cooperative localization, sensor networks, ultra wideband, measurement and modeling, cooperative localization algorithms. Includes bibliographical references (leaves 137-148).

This thesis is focused on collecting data from commercial geolocation databases and its analysis. The theoretical part discusses techniques of IP geolocation, briefly describes commercial geolocation databases and mechanisms for accessing their geolocation services. The practical part of this thesis involves implementation of scripts for collecting and modification data retrieved from commercial geolocation databases for further analysis. First script collects the data and store them in output files in specified format. Second script modifies gathered data to achieve better accuracy. Detailed analysis is performed afterwards. Analysis is performed at several levels. This includes country, region and city accuracy as well as accuracy on latitude and longitude parameters. Output of this analysis will be conclusive, and will also display accuracy of tested geolocation services. The aim of this thesis is to analyze currently available commercial geolocation services and to prove their accuracy. Used scripts are implemented in Python programming language in version 3 using modules urllib, re and json. Scripts are test on commodity hardware with Linux operating system.

The Software Communications Architecture (SCA) specification defines a framework that allows modular software components to be developed and assembled to build larger radio applications. The specification allows for these components to be distributed among a set of computing hardware and to be connected by standard interfaces. This research aims to build a spatially distributed SCA application for the Open Source SCA Implementation: Embedded (OSSIE) implementation using low-cost Universal Software Radio Peripheral (USRP) hardware. The system collects signals from multiple spatially distributed collection devices and use those signals to compute precision estimates for the location of emitters using time difference of arrival (TDOA) computations. Several OSSIE components and tools are developed to support this research. Results are presented showing the capabilities of the geolocation system. / Master of Science

ITC/USA 2015 Conference Proceedings / The Fifty-First Annual International Telemetering Conference and Technical Exhibition / October 26-29, 2015 / Bally's Hotel & Convention Center, Las Vegas, NV / Tracking small marmosets over a large area represents a significant challenge for researchers. The native habitat for such animals are generally unsuitable for GPS based location systems, and the size of the animals prevents large, feature-rich collars from being utilized. Additionally, costs and feasibility prevent researchers from continuously monitoring these animals on the ground. This paper proposes a new system of tracking that offsets complexity from the collar onto fixed Base Stations (BS). The simplified collars emit a ping that multiple BSs then log along with the power of the signal. Combining the data from different BSs allows for the determination of the Signal of Interest (SOI). It was found that using three BSs provided enough accuracy to determine the location of an SOI within an accuracy of 2 m² over a roughly 450 m² area.

Recently, a number of empirical models have been introduced in the literature for the behavior of direct path used in the design of algorithms for RF based indoor geolocation. Frequent absence of direct path has been a major burden on the performance of these algorithms directing researchers to discover algorithms using multipath diversity. However, there is no reliable model for the behavior of multipath components pertinent to precise indoor geolocation. In this dissertation, we first examine the absence of direct path by statistical analysis of empirical data. Then we show how the concept of path persistency can be exploited to obtain accurate ranging using multipath diversity. We analyze the effects of building architecture on the multipath structure by demonstrating the effects of wall length and wall density on the path persistency. Finally, we introduce a comprehensive model for the spatial behavior of multipath components. We use statistical analysis of empirical data obtained by a measurement calibrated ray-tracing tool to model the time-of- arrival, angle-of-arrival and path gains. The relationship between the transmitter-receiver separation and the number of paths are also incorporated in our model. In addition, principles of ray optics are applied to explain the spatial evolution of path gains, time-of-arrival and angle-of-arrival of individual multipath components as a mobile terminal moves inside a typical indoor environment. We also use statistical modeling for the persistency and birth/death rate of the paths.

A identificação de focos de transmissão pode ser de grande utilidade no controle da malária. Por esse motivo, hospitais em regiões endêmicas buscam saber os locais que foram visitados anteriormente por pacientes. No entanto, tais informações, obtidas através de questionários fornecidos aos pacientes, são geralmente vagas e muitas vezes imprecisas. Isto torna o processo manual, lento e de pouca valia em estudos epidemiológicos de larga escala. Baseando-se no fato de que uma parcela significativa da população possui celulares com GPS, o objetivo deste projeto é melhorar a acurácia, organização e dinâmica do processo de coleta de dados de geolocalização de pacientes infectados. Um sistema (https://sipos.fcf.usp.br) foi desenvolvido para que pacientes que chegam aos hospitais possam, sob consentimento voluntário, fornecer os dados de GPS dos seus celulares. Os dados dos usuários, que são tratados de forma anônima, são automaticamente processados e armazenados de forma segura. Através do sistema SiPoS Explorer, epidemiologistas e especialistas em saúde pública podem explorar e analisar os dados de geolocalização, permitindo, desta forma, que regiões vulneráveis sejam priorizadas durante campanhas de controle. / The identification of regions with high rates of infection can be of great use in the control of malaria. For this reason, hospitals in endemic regions seek to know the places previously visited by patients. However, such information, obtained through questionnaires provided to patients, is usually vague, inaccurate and not integrated into databases. This makes the process manual, slow and of little value in large-scale epidemiological studies. Based on the fact that a significant portion of the population has smartphones equipped with GPS, this project aims to improve the accuracy and organization of the process of collecting geolocation data from infected patients. The Sickness Positioning System (https://sipos.fcf.usp.br) was developed so that patients who arrive at hospitals can, with voluntary consent, provide the GPS data collected by their smartphones. User data, which is handled anonymously, is automatically processed and securely stored. Through the SiPoS Explorer system (https://sipos.fcf.usp.br/explorer), epidemiologists and public health experts can explore and analyze geolocation data, thereby allowing vulnerable regions to be prioritized during control campaigns.