Dissertation Defense: Zhongyuan Zhao
"Improving Spectrum Efficiency by Exploiting User and Channel Behaviors for Next Generation Wireless Networks"
Committee Members: Dr. Mehmet Can Vuran (Advisor), Dr. Massimiliano Pierobon, Dr. Stephen Scott, Dr. Lisong Xu and Dr. Demet Batur
Thursday, April 4, 2019
9 a.m.
211 Schorr Center
Abstract: The spectrum of wireless networks has been overcrowded by proliferating wireless devices under conventional license-based spectrum access policies. Dynamic Spectrum Access (DSA) policies have recently been adopted within recent regulatory models such as the TV white space (TVWS) and Citizens Broadband Radio Spectrum (CBRS) standards. However, due to the limited white spaces in many populated metropolitan and coastal cities, the study of new spectrum sharing approaches based on interactions among licensees and secondary users is needed for more efficient and fair spectrum utilization.
In this dissertation, the behaviors of users and the wireless channel are incorporated into the design of DSA solutions. New spectrum opportunities in populated areas are identified and characterized under the proposed framework of TV set-facilitated spectrum sharing. The aggregate interference in this new paradigm is modeled to facilitate spectrum management. Moreover, dynamic pricing solutions, which were developed to address operational challenges of wireless service providers under stochastically changing capacity and demand, are evaluated under realistic simulation scenarios to validate the theoretical assumptions. Furthermore, deep neural network-based physical layer is explored to enhance wireless communication in adverse radio environments under interactive spectrum sharing. Finally, to facilitate real-world experimentation of DSA solutions, in collaboration with the City of Lincoln, a city-wide testbed is developed to improve the accessibility of experimental research for the next generation wireless networks.
This thesis provides a holistic approach to improve spectrum utilization efficiency to improve existing DSA policies. Meanwhile, it fosters further discussions on technical, policy, and operational aspects of DSA solutions, and accelerates experimental researches for next generation wireless networks.
Dissertation Defense: Shruti Daggumati
“Visual Similarity Analysis of Ancient Scripts”
Monday, April 8, 2019
10:30 a.m.
103C Avery Hall
Committee Members: Dr. Peter Revesz (Advisor), Dr. Mohammad Hasan, Dr. Hongfeng Yu, and Dr. Robert Gorman
Abstract: Archaeologists and researchers have studied the relationship among ancient scripts since their discovery. The vast image libraries which have recently become available facilitate the computational study of the evolution of these ancient scripts. In this research, we see how various scripts may be classified to belong to particular script families. We focus on the Indus Valley Script, which is a completely undeciphered script and whose origin is highly debated.
This dissertation makes three major contributions. The first contribution is to the analysis of ancient script similarities using deep learning methods. We designed an algorithmic solution solely based on computer recognition of the scripts; thereby eliminating human bias. Our programs can take as input script image sets and classify them using their similarity to standard symbols. The second contribution is to the statistical analysis and refinement of the Indus script. In the script set of around seven hundred, we were able to reduce the sample size by ten percent. In addition, we algorithmically remove symbols which were once thought to be different entities. Numerous scholars have debated over the script size and whether to scale the symbol set down. However, none have used an algorithmic process to remove and reduce the dataset. The third contribution of this dissertation is to the hierarchical tree showing the similarity between the ancient scripts. We also create a plausible hierarchical tree to indicate the evolutionary relationship of the ancient scripts.
Master’s Thesis Defense: Qi Xia
“Feasibility and Security Analysis of Wideband Ultrasonic Radio for Smart Home Applications”
Wednesday, April 10, 2019
11 a.m.
211 Schorr Center
Committee Members: Dr. Qiben Yan (Advisor), Dr. Hongfeng Yu and Dr. Lisong Xu
Abstract: Smart home Internet-of-Things (IoT) accompanied with smart home apps has witnessed a tremendous growth in the past few years. Yet, smart home IoT devices as well as their apps’ security and privacy have raised serious concerns, as they are getting increasingly complicated each day, expected to store and exchange extremely sensitive personal data, always on and connected, and commonly exposed to any users in a sensitive environment. Nowadays wireless smart home IoT devices rely on electromagnetic wave-based radio-frequency (RF) technology to establish fast and good quality network connections. However, RF has its limitations that can negatively affect the smart home user experience and even cause serious security issue, such as crowded spectrum resources and wide-spreading RF waves. To overcome those limitations, people have to use technology for sophisticated time and frequency division management and rely on the assumptions that the attackers have limited computational power, which is resource consuming and insecure to powerful attackers. In this paper we proposed URadio, a wideband ultrasonic communication system, using electrostatic ultrasonic transducers. We design and develop two different types of transducer membranes using two types of extremely thin materials, Aluminized Mylar Film (AMF) and reduced Graphene Oxide (rGO), for assembling transducers, which achieve at least 45 times more bandwidth than commercial transducers. Equipped with the new wideband transducers, an OFDM communication system is deigned to better utilize the available 600 kHz wide bandwidth. Our experiments show that URadio can achieve unprecedentedly 4.8 Mbps data rate with a communication range of 17 cm. The attainable communication range is increased to 31 cm and 35 cm with data rates of 1.2 Mbps and 0.6 Mbps using QPSK and BPSK, respectively. Although the current wideband system only supports short range communication, it is expected to extend the trasmission range with better acoustic engineering. Also, by doing experiments of measuring the ultrasonic adversaries’ eavesdropping and jamming performance, we proved that our system is physically secure even when exchanging plaintext data.
Master’s Thesis Defense: Qicheng Lin
Wednesday, April 10, 2019
3:00 p.m.
112 Schorr Center
“User Privacy Leakage in Location-Based Mobile Ad Service”
Committee Members: Dr. Qiben Yan, Advisor, Dr. Hau Chan, and Dr. Witawas Srisa-an
Abstract: The online advertising ecosystem leverages its massive data collection capability to learn the properties of users for targeted ad deliveries. Many Android app developers include ad libraries in their apps as a way of monetization. These ad libraries contain advertisements from the sell-side platforms, which collect an extensive set of sensitive information to provide more relevant advertisements for their customers. Existing efforts have investigated the increasingly pervasive private data collection of mobile ad networks over time. However, there lacks a measurement study to evaluate the scale of privacy leakage of ad networks across different geographical areas. In this work, we present a measurement study of the potential privacy leakage in mobile advertising services conducted across different locations. We develop an automated measurement system to intercept mobile traffic at different locations and perform data analysis to pinpoint data collection behaviors of ad networks at both the app-level and organization-level. With 1,100 popular apps running across 10 different locations, we perform extensive threat assessments for different ad networks. Meanwhile, we explore the ad-blockers’ behavior in the ecosystem of ad networks, and whether those ad-blockers are actually capturing the users’ private data in the meantime of blocking the ads. We find that: the number of location-based ads tends to be positively related to the population density of locations, ad networks collect different types of data across different locations, most ad networks are capable of collecting precise location data, and ad-blockers can effectively block the private data leaked to the ad networks.