Defense Notices

Amazon.com is one of the largest online retail stores in the world. Besides selling millions of product on their website, Amazon provides a variety of Web services including Amazon Review and Recommendation System. Users are encouraged to write product reviews to help others to understand products’ features and make purchase decisions. However, product reviews, as a type of user generated content (UGC), suffer from quality and trust problems. To help evaluating the quality of reviews, Amazon also provides the users with the helpfulness vote feature so that a user can support a review that he considers helpful. In this project we aim to study the relation between helpfulness votes and the ranks of the reviews. In particular, we are looking for answers to questions such as “how does the helpfulness votes affect review ranks?” and “how review rank and its presentation mechanism affect people’s voting behavior?” To investigate on these questions, we built a crawler to collect reviews and votes of reviews from Amazon at a daily basis. Then, we conducted an analysis on a dataset with over 50,000 Amazon reviews to identify the voting patterns and their impact on the review ranks. Our results show that there exists a positive correlation between the review ranks and the helpfulness votes.​

Rough set theory is a popular approach for decision rule induction. However, it requires the objects in the information system to be completely described. Many real life data sets are incomplete, so we cannot apply directly rough set theory for rule induction. This project implements and compares two generalizations of rough set theory, used for rule induction from incomplete data: Tolerance Relation and Valued Tolerance Relation. A comparative analysis is conducted for the lower and upper approximations and decision rules induced by the two methods. Our experiments show that Valued Tolerance Relation provides better approximations than Simple Tolerance Relation when the percentage of missing attribute values in the datasets is high.

Rough set theory is a popular approach for decision rule induction. However, it requires the objects in the information system to be completely described. Many real life data sets are incomplete, so we cannot apply directly rough set theory for rule induction. This project implements and compares two generalizations of rough set theory, used for rule induction from incomplete data: Tolerance Relation and Valued Tolerance Relation. A comparative analysis is conducted for the lower and upper approximations and decision rules induced by the two methods. Our experiments show that Valued Tolerance Relation provides better approximations than Simple Tolerance Relation when the percentage of missing attribute values in the datasets is high.

Multi-agent systems are a common paradigm for building distributed systems in different domains such as networking, health care, swarm sensing, robotics, and transportation. Systems are usually designed or adjusted in order to reflect the performance trade-offs made according to the characteristics of the mission requirement. 
Research has acknowledged the crucial role that communication plays in solving many performance problems. Conversely, research efforts that address communication decisions are usually designed and evaluated with respect to a single predetermined performance goal. This work introduces Goal-Driven Communication, where communication in a multi-agent system is determined according to flexible performance goals. 
This work proposes an evolutionary approach that, given a performance goal, produces a communication strategy that can improve a multi-agent system’s performance with respect to the desired goal. The evolved strategy determines what, when, and to whom the agents communicate. The proposed approach further enables tuning the trade-off between the performance goal and communication cost, to produce a strategy that achieves a good balance between the two objectives, according the system designer’s needs. 

Communication systems laboratory is a hands-on way to effectively visualize the real life applications of communication systems in its simplest form. Recently, hardware equipment such as spectrum analyzer, oscilloscope, and function generator were replaced by Pico Scope 6, a software based data analyzer. The Pico Scope 6 is a user friendly software which enables its users to capture and analyze analog and digital signals with a comparatively higher accuracy. Additionally, it is an economically viable solution, from both the procurement and maintenance stand point. The current effort focuses on developing a laboratory user manual, based on Pico Scope 6, for undergraduates of the Department of Electrical Engineering and Computer Science (EECS). The series of laboratory exercises developed follows the course outline of Introduction to Communication Systems – EECS 562. The expected outcomes of this laboratory manual is an improved understanding of analog modulations, digital modulations, and noise analysis of communication systems.

The operational amplifier is perhaps the most useful integrated device in existence today. It is widely used in analogue computers simulation systems and in a variety of electronic applications such as amplification filtering, buffering and comparison of signed levels. In this design project, we use the operational amplifier for amplification. Two-stage opamp is one of the most commonly used opamp architectures. A two stage differential amplifier is designed with an objective of a minimum gain of 65 dB. The gain achieved is 74.6 dB and 71.4 MHz 3dB gain bandwidth, which is useful for medium frequency operations. The schematic circuit is constructed using Metal Oxide Semiconductor Field Effect Transistor and the technology used for the final layout is Complementary metal–oxide–semiconductor (CMOS) using Cadence.

Software Defined Networking (SDN) decouples the control or logical plane of a network from its physical/data plane thus enabling features such as centralized control, network programmability, virtualization, network application development, automation and more. However, SDN is still vulnerable to attacks and failures just like any other non-SDN network. The failure in SDN can be either a link or device failure. Controller is the central device, acting like the brain of a network, and its failure can propagate rapidly rendering the underlying data plane dysfunctional. The concept of Multi-Controller SDN uses redundancy as an effective method to ensure resilience and fault-tolerance in a Software-Defined Network. Multiple Controllers are connected in a cluster to form a physically distributed but logically centralized network. The backup controllers ensure resilience against failure, attack, disaster and other network disruptions. In this project, we implement multi-controller SDN and measure performance metrics such as high availability, reliability, latency, datastore persistency and failure recovery time in a clustered environment.

Three-dimensional (3D) integrated circuits (ICs) offer a promising near-term solution for pushing beyond Moore’s Law because of their compatibility with current technology. Through silicon vias (TSVs) provide electrical connections that pass vertically through wafers or dies to generate high-performance interconnects, which allows for higher design densities through shortened connection lengths. In recent years, we have seen tremendous technological and economic progress in adoption of 3D ICs with TSVs for mainstream commercial use. 
Along with the need for low-cost and high-yield process technology, the successful application of TSV technology requires further optimization of the TSV electrical modeling and design. In the millimeter wave (mmW) frequency range, the root mean square (rms) height of the through silicon via (TSV) sidewall roughness is comparable to the skin depth and hence becomes a critical factor for TSV modeling and analysis. The impact of TSV sidewall roughness on electrical performance, such as the loss and impedance alteration in the mmW frequency range, is examined and analyzed. The second order small analytical perturbation method is applied to obtain a simple closed-form expression for the power absorption enhancement factor of the TSV. In this study, we propose an accurate and efficient electrical model for TSVs which considers the TSV sidewall roughness effect, the skin effect, and the metal oxide semiconductor (MOS) effect. The accuracy of the model is validated through a comparison of circuit model behavior for full wave electromagnetic field simulations up to 100 GHz. 
Another advanced neurophysiological computing system that can incorporate 3D integration could provide massive parallelism with fast and energy efficient links. While the 3D neuro-inspired system offers a fantastic level of integration, it becomes inordinately arduous for the designer to model, merely because of the innumerable interconnected elements. When a TSV array is utilized in a 3D neuromorphic system, crosstalk has a malefic effect upon the system’s signal to noise ratio; the result is an overall deterioration of system performance. To countervail the crosstalk, we propose a novel optimized TSV array pattern by applying the force directed optimization algorithm. 

As our world becomes more connected, the average person must place more trust in cloud systems for everyday transactions. We rely on banks and credit card services to protect our money, hospitals to conceal and selectively disclose sensitive health information, and government agencies to protect our identity and uphold national security interests. However, establishing trust in remote systems is not a trivial task, especially in the diverse, distributed ecosystem of todays networked computers. Remote Attestation is a mechanism for establishing trust in a remotely running system where an appraiser requests information from a target that can be used to evaluate its operational state. The target responds with evidence providing configuration information, run-time measurements, and authenticity meta-evidence used by the appraiser to determine if it trusts the target system. For Remote Attestation to be applied broadly, we must have attestation protocols that perform operations on a collection of applications, each of which must be measured differently. Verifying that these protocols behave as expected and accomplish their diverse attestation goals is a unique challenge. An important first step is to understand the structural properties and execution patterns they share. In this thesis I present a semantic framework for attestation protocol execution within the Coq verification environment including a protocol representation based on Session Types, a dependently typed model of perfect cryptography, and an operational execution semantics. The expressive power of dependent types constrains the structure of protocols and supports precise claims about their behavior. If we view attestation protocols as programming language expressions, we can borrow from standard language semantics techniques to model their execution. The proof framework ensures desirable properties of protocol execution, such as progress and termination, that hold for all protocols. It also ensures properties of authenticity and secrecy for individual protocols.

The Internet of Things (IoT) is expected to revolutionize the world through its myriad applications in health-care, public safety, environmental management, vehicular networks, industrial automation, etc. Some of the concepts related to IoT include Machine Type Communications (MTC), Low power Wireless Personal Area Networks (LoWPAN), wireless sensor networks (WSN) and Radio-Frequency Identification (RFID). Characterized by large amount of traffic with smart decision making with little or no human interaction, these different networks pose a set of challenges, among which security, energy, reliability and latency are the most important ones. First, the open wireless medium and the distributed nature of the system introduce eavesdropping, data fabrication and privacy violation threats. Second, the large number of IoT devices are expected to operate in a self-sustainable and self-sufficient manner without degrading system performance. That means energy efficiency is critical to prolong devices' lifetime. Third, many IoT applications require the information to be successfully transmitted in a reliable and timely manner, such as emergency response and health-care scenarios. To address these challenges, we propose low-complexity approaches by exploiting the physical layer and using stochastic geometry as a powerful tool to accurately model the spatial locations of ''things''. This helps provide a tractable analytical framework to provide solutions for the mentioned challenges of IoT.