Volgenau School of Engineering Graduate Research

Permanent URI for this collection

https://hdl.handle.net/1920/11421

Browse

Browsing Volgenau School of Engineering Graduate Research by Issue Date

Filter results by year or month
Now showing 1 - 20 of 72
  • Thumbnail Image
    Item
  • Thumbnail Image
    Item
    Survey on Performance Analysis of Virtualized Systems
    (2019) Wang, Han
    This survey goes over the important concepts in virtualization in the overview section, after that we cover the virtualization implementations, performance analysis challenges, and virtualization measurement tool.
  • Thumbnail Image
    Item
    Multicore Processor Performance Analysis
    (2019) Kolhe, Gaurav S.
    Central Process Units (CPUs) are becoming the standard processors of current computing systems design. With the increasing performance requirements, the number of transistors on single chip unit cannot grow exponentially due to the limit area, power, and heat dissipation, etc. Therefore, multicore processor design has become a trend for current processor design. A multicore processor is an integrated computing component composed of several (two or more) CPU cores that can execute the program instructions. The individual cores can execute multiple instructions in parallel, thus significantly increasing the performance of programs which takes advantage of the unique architecture. In this survey, we introduce several aspects to demonstrate a thorough survey for multicore processor, including (1) The Need for Multicore CPU; (2) The Need for Performance Analysis; (3) The Ways of Evaluating multicore CPU Performance; (4) Factors that Affects the Performance; and (5) Multicore Benchmarking. Finally, we will discuss the existed problems, as well as future directions of multicore processor design and give the conclusions of our multicore performance analysis survey.
  • Thumbnail Image
    Item
    Cloud Computing: Literature Review
    (2019) Hassan, Rakibul
    Cloud computing has recently emerged as a new paradigm for hosting and delivering services over the Internet. Cloud computing is attractive to business owners as it eliminates the requirement for users to plan ahead for provisioning, and allows enterprises to start from the small and increase resources only when there is a rise in service demand. However, despite the fact that cloud computing offers huge opportunities to the IT industry, the development of cloud computing technology is currently at its infancy, with many issues still to be addressed. In this paper, we present a survey of cloud computing, highlighting its key concepts, architectural principles, state-of-the-art implementation as well as research challenges. The aim of this paper is to provide a better understanding of the design challenges of cloud computing and identify important research directions in this increasingly important area.
  • Thumbnail Image
    Item
    Optogenetics: Using Light to Investigate and Potentially Treat Neurological and Psychological Disorders
    (2019) Kub, Michael
    Optogenetics is an innovative neuromodulation technique involving the use of light and light-sensitive proteins to control molecular events within a genetically modified cell. The fundamental mechanism behind optogenetics is the deliberate shining of light at light-sensitive cellular membrane proteins which causes some sort of change within a cell. These proteins, called opsins, come in many forms including ion channels, pumps, and Gprotein-coupled receptors and they are found in a wide range of organisms from vertebrates to prokaryotes. When utilizing optogenetics, researchers must make several considerations including the light source to be used to control the cellular event, the type of cell to be activated by the light,and the tools to be utilized for measuring such cellular activity. We reviewed in detail the mechanism behind optogenetics and the considerations researchers make in employing this technique. We also reviewed outcomes from several studies centered around it and its current limitations. In conducting this review, we utilized web-based archives such as PubMed, Nature, and ScienceDirect. The studies that we specifically reviewed include the application of optogenetics for analyzing the effect that grafted cells have on relieving Parkinson’s Disease symptoms in animal models, the capability of optogenetics in instantly controlling depression-like states in mice, and the capability of optogenetics in regulating epilepsy in cultured animal brain models. In each of these studies, the type of cell that was sought to be controlled was the neuron, which all studies had substantial success in doing so. One area which was not addressed in these studies and which should be in future studies, is the plausibility that optogenetics could someday be used on humans. Based on the outcomes of these studies and the overall indication that optogenetics is an effective and precise technique in evoking cellular events, we conclude that optogenetics will likely have an enormous impact on research for years to come. Furthermore, given concerns over safety and use on humans, which we get into later in this paper, we also conclude that optogenetics has an uncertain future for clinical application.
  • Thumbnail Image
    Item
    Technical Report on Logic Obfuscation for Protecting the Semiconductor Intellectual Properties in the Manufacturing Supply Chain
    (2019) Roshanisefat, Shervin
    The increasing cost of building, operating, managing, and maintaining state-of-the-art silicon manufacturing facilities has pushed several stages of the semiconductor device’s manufacturing supply chain offshore. However, many of these offshore facilities are identified as untrusted entities. Processing and fabrication of ICs in an untrusted supply chain pose a number of challenging security threats such as IC overproduction, Trojan insertion, Reverse Engineering, Intellectual Property (IP) theft, and counterfeiting. To counter these threats, various hardware design-for-trust techniques have been pro-posed. Logic obfuscation, as a proactive technique among these techniques, has been introduced as a technique that obfuscates and conceals the functionality of IC/IP using additional key inputs that are driven by an on-chip tamper-proof memory. Shortly after introducing the primitive logic locking solutions, a very strong attack based on the satisfiability solvers (SAT) was shown that could break all previously proposed locking mechanisms in almost polynomial time. To thwart this attack, researchers have investigated many directions, such as formulating locking solutions that significantly increase the number of required SAT iterations, or formulating the locking solutions such that it is not translatable to a SAT problem. However, further investigations demonstrated that some of these locking techniques are vulnerable to other types of attacks such as Signal Probability Skew (SPS) attack, removal attack, approximate attacks, bypass attack, and Satisfiability Module Theories (SMT) attack. Besides, these techniques suffer from very low output corruption. Hence, an unactivated IC behaves almost identical to an unlocked IC except one or a few inputs. In this report, first, we will characterize the SAT attack, which shows that how using different SAT solvers can produce different results with large deviations which demonstrates that long execution time or high memory usage in one SAT solver may not be a problem in another solver. Next, we discuss a branch of SAT-resilient methods called cyclic locking and propose efficient methods to introduce feedbacks into a circuit in a way that SAT and its improved versions for cyclic circuits could not find the correct key. Then, we discuss a new branch of obfuscation techniques that tries to restrict access to the scan chain and thus circumvent the SAT attack. In there, we discuss a new attack method called unrolling SAT that potentially could be used for breaking obfuscated scan chains and recover the protected design.
  • Thumbnail Image
    Item
    A Technical Report on Logic Obfuscation using Reconfigurable Logic and Routing Blocks
    (2019) Mardani Kamali, Hadi
    The increasing cost of building, operating, managing, and maintaining state-of-the-art silicon manufacturing facilities has pushed several stages of the semiconductor device’s manufacturing supply chain offshore. However, many of these offshore facilities are identified as untrusted entities. Processing and fabrication of ICs in an untrusted supply chain poses a number of challenging security threats such as IC overproduction, Trojan insertion, Reverse Engineering, Intellectual Property (IP) theft, and counterfeiting. To counter these threats, various hardware design-for-trust techniques have been pro- posed. Logic locking, as a proactive technique among these techniques, has been introduced as a technique that obfuscates and conceals the functionality of IC/IP using additional key inputs that are driven by an on-chip tamper-proof memory. Shortly after introducing the primitive logic locking solutions, a very strong Boolean attack, the Satisfiability (SAT) attack. It was shown that the SAT attack could break all previously proposed primitive locking mechanisms in almost polynomial time. To thwart the strength of SAT attack, researchers have investigated many directions, such as formulating locking solutions that significantly increase the number of required SAT iterations, or formulating the locking solutions such that it is not translatable to a SAT problem. However, further investigations demonstrated that some of these locking techniques are vulnerable to other types of attacks such as Signal Probability Skew (SPS) attack, removal attack, approximate-based attack(s), bypass attack, and Satisfiability Module Theories (SMT) attack. In addition, these techniques suffer from very low output corruption. Hence, an unactivated IC behaves almost identical to an unlocked IC with exception of one or few inputs.
  • Thumbnail Image
    Item
  • Thumbnail Image
    Item
    Beyond Bibliometrics: Understanding Library Services in Multidisciplinary Research
    (Mason Graduate Interdisciplinary Conference, 2019-04-06) Mahapasuthanon, Pattiya; Hoffman, Kimberly
    Bibliometric methods, using citations as data, are an alternate way to learn from the literature of science and technology. These statistical methods are used, with visualization tools, to determine the relationship between authors and papers, scientific subjects reflected in publishing, and word or frequency occurrence. Bibliometrics are used by libraries to get a broad view of the growth, structure, connections and productivity of a discipline reflected in literature. This research studies trends and multidisciplinary connections across university researchers and campuses. With a strategic initiative from George Mason University (GMU) to become excellent in multidisciplinary research, Mason Libraries support multidisciplinary research activities. This study completes an analysis on bibliometric and funding across five centers at the GMU Science and Technology campus (SciTech) to understand research activities and interactions. Bibliometric network graphs were created from Web of Science (WoS) citation datasets and VOSViewer, a visualization tool. Federal RePORTER [documentation and analysis of inputs, outputs, and outcomes resulting from federal investments in science available: https://www.starmetrics.nih.gov/Star/About] and WoS were used to generate funding charts. For research activities not captured by scientific literature, and involving ongoing library resources, the creation of a pilot version of an interactive visualization for experience mapping was tested to successfully identify and seek new service opportunities. The results obtained from the bibliometric analysis indicate that libraries must plan to reach researchers in those relatively young multidisciplinary research institutes. The research trends at SciTech have shifted towards applied health and biological medicine according to the keyword analysis. From the funding analysis, the SciTech campus accounted for 30 percent of the total funded projects to GMU from National Institutes of Health (NIH). With these preliminary results, understanding resources and services the SciTech researchers and multidisciplinary researchers need will increase research connections and productivity. Future research will seek to incorporate more sophisticated tools to further understand impactful resources and plan for future library collaborations.
  • Thumbnail Image
    Item
    A Survey On Techniques Used for Designing Fault Tolerant and Process Variation Aware Memories and Caches
    (2020) Beheshti-Shirazi, Sayed Aresh
    Aggressive voltage and frequency scaling schemes applied to memory and cache structures, specially for memory systems fabricated in advanced and scaled geometry nodes that are severely affected by process variation,significantly increases the likelihood of read, write and access failures to/from memory cell array, and reduces the extent of frequency and voltage scaling. To remedy this problem, in the past decade, many researchers have investigated alternative and fault tolerant cache and memory organizations to mitigate the impact of process variation, and to reduce the failure rate of memory array in the results of voltage and frequency scaling. This paper discusses and compare many of such cache and memory design techniques.
  • Thumbnail Image
    Item
  • Thumbnail Image
    Item
    A Side Channel Delay Analysis for Hardware Trojan Detection
    (2020-10) Vakil, Ashkan
    This research proposal introduces a learning assisted modeling technique for the purpose of Hardware Trojan detection. Our proposed model, unlike the prior art, does not require a Golden fabricated chip as a fingerprint to compare the side channel signals. Instead, by modeling the voltage drop and voltage noise pre-fabrication, and with training a Neural Network post-fabrication, our proposed technique can improve the timing model collected during timing closure and produces a Neural assisted Golden Timing Model (NGTM) for side channel delay-signal analysis. The Neural Network acts as a process tracking watchdog for correlating the static timing data (produced at design time) to the delay information obtained from clock frequency sweeping test. Proposed modeling technique enables Hardware Trojan detection close to 90% in the simulated scenarios.
  • Thumbnail Image
    Item
    A CNN/MLP Neural Processing Engine, Powered by Novel Temporal-Carry-deferring MACs
    (2021) Mirzaeian, Ali
    The applications of machine learning algorithms are innumerable and cover nearly every domain of modern technology. During this rapid growth of this area, more and more companies have expressed a desire to utilize machine learning techniques in smaller devices, such as cell phones or smart Internet of Things (IoT) instruments. However, as machine learning has so far required a power source with more capacity and higher efficiency than a conventional battery. Therefore, introducing neural network accelerators with low energy demands and low latency for executing machine learning techniques has drawn lots of attention in both the academia and industry.
  • Thumbnail Image
    Item
    Diverse Knowledge Distillation (DKD): A Solution for Improving The Robustness of Ensemble Models Against Adversarial Attacks
    (2021) Mirzaeian, Ali; Kosecka, Jana; Homayoun, Houman; Mohsenin, Tinoosh; Sasan, Avesta
    This paper proposes an ensemble learning model that is resistant to adversarial attacks. To build resilience, we introduced a training process where each member learns a radically distinct latent space. Member models are added one at a time to the ensemble. Simultaneously, the loss function is regulated by a reverse knowledge distillation, forcing the new member to learn different features and map to a latent space safely distanced from those of existing members. We assessed the security and performance of the proposed solution on image classification tasks using CIFAR10 and MNIST datasets and showed security and performance improvement compared to the state of the art defense methods.
  • Thumbnail Image
    Item
    Conditional Classification: A Solution for Computational Energy Reduction
    (2021) Mirzaeian, Ali; Manoj P D, Sai; Vakil, Ashkan; Homayoun, Houman; Sasan, Avesta
    Deep convolutional neural networks have shown high efficiency in computer visions and other applications. However, with the increase in the depth of the networks, the computational complexity is growing exponentially. In this paper, we propose a novel solution to reduce the computational complexity of convolutional neural network models used for many class image classification. Our proposed technique breaks the classification task into two steps: 1) coarse-grain classification, in which the input samples are classified among a set of hyper-classes, 2) fine-grain classification, in which the final labels are predicted among those hyper-classes detected at the first step. We illustrate that our proposed classifier can reach the level of accuracy reported by the best in class classification models with less computational complexity (Flop Count) by only activating parts of the model that are needed for the image classification.
  • Thumbnail Image
    Item
    Information Retrieval Model for Social Media Applications
    (2021-01) Bhandari, Shova
    Social networks are rich source of data to analyze user habits in all aspects of life. User’s behaviour is decisiv e component of a health system in various countries. Promoting good behaviour can improve the public health signif icantly. In this work, we develop a new model for social network analysis by using text analysis approach. We defi ne each user reaction to global pandemic with analysing his online behaviour. Clustering a group of online users w ith similar habits, help to find how virus spread in different societies. Promoting the healthy life style in the high risk online users of social media have significant effect on public health and reducing the effect of global pandemic. In this work, we introduce a new approach to clustering habits based on user activities on social media in the time of pandemic and recommend a machine learning model to promote health in the online platforms.
  • Thumbnail Image
    Item
    Machine Learning Model to Detect Emergency in the Global Pandemic
    (2021-01-11) Raju, Rini
    It is crucial to use advanced machine learning models to improve disaster and emergency response in critical events around the world. In this paper, we introduce a new model, which can highlight the essential help that people need in times of emergency. Based on the user comments, we choose the emergency response that can use the optimal resources to address the maximum needs. The new features in the model help to analyze each person's response from political, social, and health perspectives. This approach helps to recognize different types of users to improve emergency response in the time of the global pandemic. Also, collecting pandemic data from different online resources, makes this research more powerful in feature extraction to improve the model accuracy based on emergency data. This model can help health applications to improve disaster response time and services.
  • Thumbnail Image
    Item
    Sentiment Analysis Methods to Mitigate Negative Effect of the COVID-19 Pandemic
    (2021-01-11) Mohamud, Sofia A
    The goal of this research is to determine crucial factors that played a role in the number of confirmed COVID-19 infections within a given location. We hypothesize that political bias plays a significant role in the rise of COVID-19 cases globally and nationally; specifically, in overriding scientific reasoning for the delay or lack of deploying national policies to address the pandemic. Methods: To determine the validity of our hypothe- sis, we performed a literature review that identified statistical information on 1) the origins of the virus, 2) the lethality of the virus, and 3) potential parties responsible for the creation and release of the virus. In addition to the literature review, our team performed a behavioral analysis using information extracted from social media platforms to identify and determine behavior patterns associated with specific words related to the virus
  • Thumbnail Image
    Item
    Online User Profiling to Detect Social Bots on Twitter
    (2021-03) Heidari, Maryam; Jones, James H. R.; Uzuner, Ozlem
    Social media platforms can expose influential trends in many aspects of everyday life. However, the trends they represent can be contaminated by disinformation. Social bots are one of the significant sources of disinformation in social media. Social bots can pose serious cyber threats to society and public opinion. This research aims to develop machine learning models to detect bots based on the extracted user's profile from a Tweet's text. Online users' profile shows the user's personal information, such as age, gender, education, and personality. In this work, the user's profile is constructed based on the user's online posts. This work's main contribution is three-fold: First, we aim to improve bot detection through machine learning models based on the user's personal information generated by the user's online comments. The similarity of personal information when comparing two online posts makes it difficult to differentiate a bot from a human user. However, in this research, we turn personal information similarity among two online posts as an advantage for the new bot detection model. The new proposed model for bot detection creates user profiles based on personal information such as age, personality, gender, education from user's online posts, and introduces a machine learning model to detect social bots with high prediction accuracy based on personal information. Second, create a new public data set that shows the user's profile for more than 6900 Twitter accounts in the Cresci 2017\cite{cresci-etal-2017-paradigm} data set. All user's profiles are extracted from the online user's posts on Twitter. Third, for the first time, this paper uses a deep contextualized word embedding model, ELMO\cite{peters-2018-deep}, for social media bot detection task.
  • Thumbnail Image
    Item
    Machine learning models for Prediction of the need for future Covid-19 vaccine booster
    (2021-04) Marzook, Ahmad Al; Xu, Ge; Jagannath, Prajna Shetty
    About 4 million Virginia citizens are fully vaccinated against COVID-19. According to medical data from nations like Israel and the United Kingdom, all those people may require another shot in the near future, at least when keeping transmis- sible diseases at bay. Health officials are looking at whether a booster shot is needed to ensure the vaccine’s effectiveness. To prepare it for predictions, we will develop a new simple logistic regression model. Make prediction models using classification and probability. With the vaccine immune response lifetime taken into account, we will run our model to forecast the completely vaccinated number timeline and observe when it achieves the herd immunity percentage of the overall Virginia population. The percentage of people that need to be vaccinated to attain COVID- 19 herd immunity is still up for dispute. The goal of this study was to forecast when everyone would be fully immunized. COVID-19 vaccination loses nearly half of its defense antibodies every 108 days, according to Elie Dolgin. As a result, immunizations that initially provided 90% protection against mild episodes of the disease may only provide 70% protection after 6 or 7 months. The percentage of people that need to be vaccinated to attain COVID-19 herd immunity is still being disputed.