教学课程资源库 更多>>

关键词： Communication   Electrical engineering   Computer science  

摘要： Telecommunication systems have been through continuous evolution to keep up the fast-growing network traffic demand. With developments such as HD video streaming, cloud computing, Internet of Things (IoT), hyper-scale data centers, and 5G wireless networks, more demanding network requirements raise challenges to create a more efficient optical communication system that can provide the capability to support a wide range of applications. Specifically, 5G standards require more bandwidth and ultra-low latency; metro-scale optical aggregation networks motivate more scalable and on-demand optical network capacity. Dynamic reconfigurable optical add-drop multiplexer (ROADM) based wavelength-division multiplexing (WDM) systems in which connections are established through real-time wavelength switching operations have long been studied as a means of achieving greater scalability and increasing the network resource utilization. A new dimension referred to as disaggregated optical systems, has the potential to further drive down cost by commoditizing the hardware. While ROADMs are extensively deployed in today’s WDM systems, their interoperability and functionality remain limited. Recent advances in hardware and software such as optical physical layer software-defined networking (SDN) significantly improve the multi-layer control and management potential of ROADM systems even facilitating wavelength switching. However, ensuring stable performance and reliable quality of transmission (QoT) remain severe problems, particularly for disaggregated systems. A key challenge in enabling disaggregated optical systems is the uncertainty and optical power dynamics that arise from a variety of physical effects in the amplifiers and transmission fiber. This thesis examines the potential for machine learning for QoT estimation in software defined networking control, and its application to disaggregated ROADM systems. Current physical layer control of flexible meshed optical networks with d

关键词： Device-to-device communication   Hidden Markov models   Tutorials   Computer science   Analytical models   Steady-state   Frequency measurement   Device-to-device communication   mobility   mobile communication   mobile data traffic  

摘要： Device-to-device (D2D) communication proposes a new epitome in mobile networking to avail data exchange between physically proximate devices. The exploitation of D2D communication enables mobile operators to harvest short range communications for improving network performance and corroborating proximity-based services. In this paper, we investigate mobility aspects of D2D communication, which are indispensable for the adoption and implementation of D2D communication technology. We present an extensive review of the state-of-the-art problems and the corresponding solutions for encouraging the exploitation of mobility to assist D2D communication. Specifically, by identifying the mobility models, traces, problems, requirements, and features of different proposals, we discuss the lessons learned and summarize the advantages of mobility-aware D2D communication. We also present open problems and highlight future research directions concerning D2D communication applications in real-life scenarios. To the best of our knowledge, this is the first comprehensive survey to address mobility-aware D2D communication, which offers insight to the underlying problems and provides the potential solutions.

关键词： Computer science  

摘要： The demand for public wireless networks has increased in recent years and is expected to continue. Available access to the Internet is an important feature that many businesses and public locations or institutions provide. With the increased use of public Wi-Fi comes an increased risk as attackers may view these networks and their users as potential targets. Wi-Fi networks are at risk due to vulnerabilities in their settings, administration, and security protocols. Attackers utilize several attack vectors to target public Wi-Fi. A significant method used by attackers is the rogue access points. These types of attacks target existing access points or disguise themselves as a legitimate device to entice users to connect. Attackers can then target the data transmissions of users connected to the rogue access point. This research identities vulnerabilities inherent to public Wi-Fi like wireless security protocols or device settings that puts the device at risk. It specifically analyzes the threat of rogue access points and how attackers use them to target public Wi-Fi. The research provides an examination of methods that administrators and users can utilize to protect identify the presence of rogue access points as well as practices that can protect user data.

关键词： Computer science  

摘要： Blockchain technology has redefined the way the software industry's core mechanisms operate. With recent generations of improvement observed in blockchain, the industry is surging ahead towards replacing the existing computing paradigms with consortium blockchain-enabled solutions. For this, there is much research observed which aims to make blockchain technology’s performance at par with existing systems. Most of the research involves the optimization of the consensus algorithms that govern the system. One of the major aspects of upcoming iterations in blockchain technology is making individual consortium blockchains collaborate with other consortium blockchains to validate operations on a common set of data shared among the systems. The traditional approach involves requiring all the organizations to run the consensus and validate the change. This approach is computationally expensive and reduces the modularity of the system. Also, the optimized consensus algorithms have their specific requirements and assumptions which if extended to all the organizations leads to a cluttered system with high magnitudes of dependencies. This thesis proposes an architecture that leverages the use of state machine replication extended to all the nodes of different organizations with seamless updates over a random graph network without involving all the nodes participating in the consensus. This also enables organizations to run their respective consensus algorithms depending on their requirements. This approach guarantees the finality of consistent data updates with reduced computations with high magnitudes of scalability and flexibility.

关键词： Information technology   Computer science  

摘要： Today’s big data world heavily relies upon providing precise, timely, and actionable intelligence, while being burdened by the ever increasing need for data cleaning and preprocessing. While in the case of ingesting large quantity of unstructured data this problem is unavoidable, when it comes to sensor networks built for a specific purpose, such as anomaly detection, some of that computation can be moved to the edge of the network. This thesis concerns the special case of sensor networks tailored for monitoring the power grid for anomalous behavior. These networks monitor power delivery infrastructure with the intent of finding deviations from the nominal steady state, across multiple geographical locations. Aforementioned deviations, known as power quality anomalies, may originate, and be localized to the location of the sensor, or may affect a sizable portion of the power grid. The difficulty of evaluating the extent of a power quality anomaly stems directly from their short temporal and variable geographical impact. I present a novel distributed power quality monitoring system called Napali which relies on extracted metrics from individual meters and their temporal locality in order to intelligently detect anomalies and extract raw data within temporal window and geographical areas of interest. The claims of this thesis are that Napali outperforms existing power quality monitoring gridwide event detection methods in resource utilization and sensitivity. Furthermore, Napali residential monitoring is capable of power grid monitoring without deployment on the high voltage transmission lines. Final claim of this thesis is that Napali capability of extracting portions of the events which did not pass the critical thresholds used in other detection methods allows for better localization of power quality disturbances. Napali claim validation was performed through deployment at the University of Hawaii. Fifteen OPQ Box devices, designed specifically to operate with Napa

关键词： Fluid mechanics   Law   Computer science  

摘要： As technology continues to progress, Constitutional protections involving technological devices are becoming the focal point in criminal proceedings. With technology evolving at a rapid rate, the courts within the United States are constantly interpreting Constitutional protections to individuals’ rights including emerging technological devices. The issue being examined herein is whether the government can force the accused to unwillingly provide a passcode to decrypt a device. The research examined a problem with the Fifth Amendment protection against self-incrimination and decryption. Furthermore, the research utilized research questions and case law study to conclude that Fifth Amendment protections do in fact, extend to those forced to provide a passcode by the government. Ultimately, the research justified the need for the utilization of cyber experts to maintain awareness of legal implications that surround the Internet of Things (IOT).

关键词： Electrical engineering   Mechanical engineering   Computer science  

摘要： This work developed strategies for reducing the computational complexity of implementing robust controllers for active magnetic bearing (AMB) systems and investigated the use of a novel add-on controller for gyroscopic effect compensation to improve achievable performance with robust controllers. AMB systems are multi-input multi-output (MIMO) systems with many interacting mechanisms that needs to fulfill conflicting performance criteria. That is why robust control techniques are a perfect application for AMB systems as they provide systematic methods to address both robustness and performance objectives. However, robust control techniques generally result in high order controllers that require high-end control hardware for implementation. Such controllers are not desirable by industrial AMB vendors since their hardware is based on embedded systems with limited bandwidths. That is why the computational cost is a major obstacle towards industry adaptation of robust controllers. Two novel strategies are developed to reduce the computational complexity of single-rate robust controllers while preserving robust performance. The first strategy identifies a dual-rate configuration of the controller for implementation. The selection of the dual-rate configuration uses the worst-case plant analysis and a novel approach that identifies the largest tolerable perturbations to the controller. The second strategy aims to redesign the controller by identifying and removing negligible channels in the context of robust performance via the largest tolerable perturbations to the controller. The developed methods are demonstrated both in simulation and experiment using three different AMB systems, where significant computational savings are achieved without degrading the performance. To improve the achievable performance with robust controllers, a novel add-on controller is developed to compensate the gyroscopic effects in flexible rotor-AMB systems via modal feedback control. The comp

关键词： Bioinformatics   Computer science  

摘要： Drug discovery is a time intensive and economically costly process. A wealth of information regarding drug targets, side effects, drug-induced gene expression signatures, among several other attributes are publicly available through existing online tools, databases, and repositories. By processing, abstracting, and aggregating information from these resources into drug set libraries, novel relationships between drugs and small molecules can be discovered. Drug set enrichment analysis leverages Fisher’s exact test to compute significant overlap between lists of drugs and small molecules generated from unbiased experiments and prior knowledge about shared attributes between small molecule sets contained in drug set libraries. Understanding the underlying molecular and structural similarities between seemingly heterogeneous sets of small molecules can aid in identifying drug repurposing opportunities, and the discovery of novel drugs. Here, we present Drugmonizome and DrugEnrichr, two web-based software applications for querying annotated sets of drugs and small molecules to match biomedical terms with input sets of drugs. We validate the usage of these tools with a case study that examines common features of published in-vitro drug screens conducted to identify inhibitors for SARS-CoV-2 as potential treatments for COVID-19.

关键词： Biomechanics   Computer science   Hammer throwing   Wearable technology   Wireless sensor networks  

摘要： Wearable sensors have been popularly used in many applications with the development of computer science and engineering. However, wearables for biomechanical feedback in motor learning and training are still rare. Therefore, this thesis focuses on developing an efficient and cost-effective wireless sensor system through a case study on the hammer throw. The results have shown that the proposed reconfigurable and wearable system can implement real-time biomechanical feedback in the hammer-throw training. Furthermore, the experimental results suggest that various throw-control patterns could be identified by using one tension-sensor and two inertial measurement units (i.e., more superior practicality than 3D motion capture), indicating that the low-cost wearable system has potential to substitute the expensive 3D motion capture technology. The proposed system can be easily modified and applied to many other applications, including but not limited to healthcare, rehabilitation, and smart homes, etc.

关键词： Operations research   Artificial intelligence   Computer science  

摘要： For many real-world automated planning problems, it is difficult to obtain a transition model that governs state evolution with complex dynamics. Fortunately, the availability of sensors allows the collection of large quantities of data which can be input to efficient machine learning algorithms to accurately approximate the unknown transition model as a neural network. The central thesis of this dissertation is that automated planning problems with unknown transition models can be efficiently learned from data using neural networks, and the resulting learned planning problems can be efficiently compiled and solved as mathematical optimization problems. In this dissertation, we formally define the deterministic factored planning problem and learn its unknown state transition function with two distinct neural network structures, namely: (i) densely-connected neural networks with Rectified Linear Units (ReLUs), and (ii) Binarized Neural Networks (BNNs). For these learned neural network structures, we provide efficient compilations based on Mixed-Integer Linear Programming (MILP), and Binary Linear Programming (BLP) and Weighted Partial Maximum Satisfiability (WP-MaxSAT), respectively. For the MILP model, we investigate different preprocessing methodologies based on reachability, network sparsification and potential functions to speed up the experimental runtime performance of the underlying MILP solver. In order to find potential functions that support efficient planning with learned neural networks, we formalize the problem of finding potential functions for learned neural networks as a Bilevel Programming model, and solve it efficiently using a finite-time constraint generation algorithm. For the WP-MaxSAT model, we introduce a novel neuron activation compilation that is asymptotically the most compact encoding in the literature that maintains generalized arc-consistency property through unit propagation -- an important property that facilitates efficiency in WP-Max