With the constant scaling of semiconductor devices, reliability of these devices is a huge concern. One of the biggest reliability issues is a phenomenon known as electromigration (EM) [1] [2]. Electromigration is the transport of material caused by the gradual movement of the ions in a conductor due to the momentum transfer between conducting electrons and diffusing metal atoms [27]. The damage induced by electromigration appears as the formation of voids and hillocks, resulting in electrical discontinuity. Based on previous Electromigration research [15], I have created a tool chain that identifies where electromigration is likely to occur in large-scale integrated circuits. Using this tool chain, it is possible to identify the mean-time to failure (MTTF) of several common and high priority circuits such as complex adders and memories. Furthermore, this tool chain allows designers to isolate weak-points in these circuits to improve the overall MTTF of the circuit. The result is that with a few simple changes, circuits can be redesigned to increase the MTTF, at minimal cost to the system.

Network reliability is the probabilistic measure that determines whether a network remains functional when its elements fail at random. Definition of functionality varies depending on the problem of interest, thus network reliability has much potential as a unifying framework to study a broad range of problems arising in complex network contexts. However, since its introduction in the 1950's, network reliability has remained more of an interesting theoretical construct than a practical tool. In large part, this is due to well-established complexity costs for both its evaluation and approximation, which has led to the classification of network reliability as a NP-Hard problem. In this dissertation we present an algorithm to estimate network reliability and then utilize it to evaluate the reliability of large networks under various descriptions of functionality. The primary goal of this dissertation is to pose network reliability as a general scheme that provides a practical and efficiently computable observable to distinguish different networks. Employing this concept, we are able to demonstrate how local structural changes can impose global consequences. We further use network reliability to assess the most critical network entities which ensure a network's reliability. We investigate each of these aspects of reliability by demonstrating some example applications. / Ph. D.

In conventional signal processing approaches, researchers need to manually extract features from raw data that can better describe the underlying problem. Such a process requires strong domain knowledge about the given problems. On the contrary, deep learning-based signal processing algorithms can discover features and patterns that would not be apparent to humans by feeding a sufficient amount of training data. In the past decade, deep learning has proved to be efficient and effective at delivering high-quality results. Deep learning has demonstrated its great advantages in image processing and text mining. One of the most promising applications of deep learning-based signal processing techniques is autonomous driving. Today, many companies are developing and testing autonomous vehicles. High-level autonomous vehicles are expected to be commercialized in the near future. Besides, deep learning has demonstrated great potential in wireless communications applications. Researchers have addressed some of the most challenging problems such as transmitter classification and modulation recognition using deep learning. Despite these advantages, there exist a wide range of security and reliability issues when applying deep learning models to real-world applications. First, deep learning models could not generate reliable results for testing data if the training data size is insufficient. Since generating training data is time consuming and resource intensive, it is important to understand the relationship between model reliability and the size of training data. Second, deep learning models could generate highly unreliable results if the testing data are significantly different from the training data, which we refer to as ``out-of-distribution (OOD)'' data. Failing to detect OOD testing data may expose serious security risks. Third, deep learning algorithms can be easily fooled when the input data are falsified. Such vulnerabilities may cause severe risks in safety-critical applications such as autonomous driving. In this dissertation, we focus on the security and reliability issues in deep learning models in the following three aspects. (1) We systematically study how the model performance changes as more training data are provided in wireless communications applications. (2) We discuss how OOD data can impact the performance of deep learning-based classification models in wireless communications applications. We propose FOOD (Feature representation for OOD detection), a unified model that can detect OOD testing data effectively and perform classifications for regular testing data simultaneously. (3) We focus on the security issues of applying deep learning algorithms to autonomous driving. We discuss the impact of Perception Error Attacks (PEAs) on LIDAR and camera and propose a countermeasure called LIFE (LIDAR and Image data Fusion for detecting perception Errors). / Doctor of Philosophy / Deep learning has provided computers and mobile devices extraordinary powers to solve challenging signal processing problems. For example, current deep learning technologies are able to improve the quality of machine translation significantly, recognize speech as accurately as human beings, and even outperform human beings in face recognition. Although deep learning has demonstrated great advantages in signal processing, it can be insecure and unreliable if the model is not trained properly or is tested under adversarial scenarios. In this dissertation, we study the following three security and reliability issues in deep learning-based signal processing methods. First, we provide insights on how the deep learning model reliability is changed as the size of training data increases. Since generating training data requires a tremendous amount of labor and financial resources, our research work could help researchers and product developers to gain insights on balancing the tradeoff between model performance and training data size. Second, we propose a novel model to detect the abnormal testing data that are significantly different from the training data. In deep learning, there is no performance guarantee when the testing data are significantly different from the training data. Failing to detect such data may cause severe security risks. Finally, we design a system to detect sensor attacks targeting autonomous vehicles. Deep learning can be easily fooled when the input sensor data are falsified. Security and safety can be enhanced significantly if the autonomous driving systems are able to figure out the falsified sensor data before making driving decisions.

This thesis investigates techniques for prototyping and evaluation of medium voltage (MV) power module packages. Specific focus will be given to the utilization of silver sintering as a bonding method for high temperature, high density power modules. Nano-silver paste and preform will be examined in detail as enabling technologies for a new generation of power electronics. To accomplish this task, analysis and characterization of the metal-ceramic substrate and its structure is performed. First, finite element models are created to evaluate the fatigue behavior of the large area bonds in the substrate structure. Prototypes of these multi-layer substrates have also been fabricated and will be subjected to thermal cycling tests for experimental verification of the efficacy of their sintered silver bonds. Stacked direct-bonded aluminum (DBA) substrates have been found to withstand up to 1000 thermal cycles of –40 °C to 200 °C when attached with low pressure-assisted silver sintering. The thermal performance of 10 kV SiC power module utilizing multi-layer DBA substrates bonded with a large-area, low pressure-assisted sintered silver bond will also be examined to ensure the sintered bond is viable for the harsh operating conditions of MV modules. A junction-to-case thermal resistance of 0.142 °C/W is measured on a module prototype utilizing stacked DBA substrates. Finally, analysis of a double-sided cooling scheme enabled by large area sintering is simulated and prototyped to demonstrate a 6.5 kV package for a MV power device. Residual stress failures induced by a highly rigid structure have been examined and mitigated through implementation of a 5 MPa pressure-assisted, double-sided silver sintering approach. / Master of Science / Power modules are the building blocks of the electrical grid of the future. As society transitions to renewable energy to fight the crisis presented by climate change, the structure of the energy grid will have to change to accommodate the increase in solar, wind, geothermal, and other renewable sources of energy generation. A clean energy grid structure will contain ubiquitous opportunities to use power modules for medium-voltage (MV) applications, like managing the flow of electricity from solar panels and wind turbines to neighborhoods and office buildings. However, these MV power modules will need to be resilient to extreme temperature and electrical stresses inherent to these applications. Current technology must be improved in both performance and reliability to match the needs of this future grid. This thesis investigates, through both experiment and computer simulation, techniques for improving the reliability of MV power modules without sacrificing thermal or electrical performance. Techniques presented in this work have the potential to transform power modules, so they may operate at higher temperatures and efficiencies for a longer lifetime than the current state-of-the-art.

Reliability is a key aspect of power system design and planning. In this research we present a reliability analysis algorithm for large scale, radially operated (with respect to substation), reconfigurable, electrical distribution systems. The algorithm takes into account equipment power handling constraints and converges in a matter of seconds on systems containing thousands of components. Linked lists of segments are employed in obtaining the rapid convergence. A power flow calculation is used to check the power handling constraints. The application of distributed generators for electrical distribution systems is a new technology. The placement of distributed generation and its effects on reliability is investigated. Previous reliability calculations have been performed for static load models and inherently make the assumption that system reliability is independent of load. The study presented here evaluates improvement in reliability over a time varying load curve. Reliability indices for load points and the overall system have been developed. A new reliability index is proposed. The new index makes it easier to locate areas where reliability needs to be improved. The usefulness of this new index is demonstrated with numerical examples. / Master of Science

Engaging in muscle strengthening activities (MSA) as part of a physical activity program offers health benefits. Although the merits of physical activity are well documented, many adults fail to meet appropriate levels as recommended in the 2008 Physical Activity Guidelines for Americans (PAGA). To get a more complete understanding on an individual's physical activity behaviors, the Tracking Resistance Exercise and Strength Training (TREST) internet based survey was developed. The purpose of the current study was to determine the test-retest reliability of TREST items. Additionally, the prevalence of participants meeting the 2008 PAGA was reported by gender. The survey was completed approximately two weeks apart by 224 (52% male) undergraduate kinesiology students. Analysis of the survey items presented TREST as a reliable instrument in assessing an individual's physical activity behavior with a focus on MSA. Among the convenience sample of 445 participants (56% male) that completed the survey in assessment #1, 73% met the 2008 PAGA minimum recommendations for MSA (>=2 days/week) and aerobic activity (>= 150 min MVPA). A more complete MSA and MVPA criteria was established (requiring MSA of all seven major muscle groups) and only 32% of participants met this guideline. In general, men engaged in aerobic exercise and MSA more than women. These results cannot be generalized due the age, activity level, and education of the study's participants. Future studies should investigate the validity of TREST items among a sample of varying fitness levels, races/ethnicities, ages, and educational levels.

The principal objective of the research is to diagnose the causes and symptoms of company failure and to investigate whether a pattern of failure could be determined to enable management and other interested parties to identify the risks threatening the survival of the company. The current research divides into three main areas 1. Development of a Data Bank and a study of the age structure of failed companies. 2. An application of reliability management techniques to the analysis of company failure data. 3. Identification of causes and symptoms of company failure based on risk management methodology. Data were collected and analysed for approximately 2000 manufacturing companies which had undergone either compulsory or creditors' voluntary liquidation during the period 1970 to 1977. A Data Bank was established with classified information for 16 different groups of companies making up the manufacturing industry. The classification was based on the Standard Industrial Classification. A study of the age structure of each group was carried out and compared with previous studies. Reliability methodology was applied to the analysis of company failure data for the identification of the failure pattern. Best distributions describing failure behaviour of companies were also determined and the validity and application of various statistical distributions were examined. A detailed examination of the histories of some large companies which failed during the period 1970-1977 was carried out. Risks, weaknesses and possible causes and symptoms of failure were investigated and discussed. A list of the causes of failure emerged from the analysis is drawn and the non-financial symptoms are highlighted in a tabular form. Illustrative models for the appraisal of change and identification of causes and symptoms are developed and critical factors discussed. Finally, general conclusions arising out of the research are-presented, along with recommendations for further research and study.

No description available.

<p> A general study of theory of reliability has been made. The mechanism of failure of mechanical components, the properties of mechanical components, and the interaction of properties responsible for failure have been investigated. Elements of probability and statistics pertinent to reliability theory have been reviewed in brief. The various testing methods for determining component reliability using exponential, normal and Weibull distributions have been investigated. Acceptance sampling procedures for satisfying the necessary reliability requirements have been presented. Use of statistical methods in predicting fatigue life of mechanical components in general and rolling contact bearings in particular have been discussed. </p> / Thesis / Master of Engineering (MEngr)

Technological innovations have lead to the development of Wearable Physiological Monitoring devices, that have enabled researchers and clinicians in real-time monitoring of physiologic function within a field setting. However, it is important to establish the psychometric properties of a device prior to its utilization. A systematic review was conducted to provide a summary and appraise the quality of the literature on psychometric parameters of Zephyr Bioharness and Fitbit devices. Based on this review, we addressed the current gaps in the literature regarding the reliability parameters of Zephyr Bioharness and Fitbit Charge devices, and established the validity and agreement properties of Fitbit Charge device. For our systematic review, we searched the Google Scholar and PubMed databases to identify articles. To establish the reliability, validity and agreement parameters of Zephyr Bioharness and Fitbit Charge devices, a convenience and snowball sampling approaches were used to recruit sixty participants (30 females) from university student, staff, faculty population, and MacSeniors Community Program at McMaster University. The performance of Zephyr and Fitbit devices were assessed throughout three phases; rest, Modified Canadian Aerobic Fitness Test and recovery. In our study, at rest, inter-session average heart rate (beats/min.) ICCs (SEM) for Zephyr and Fitbit ranged from 0.90 – 0.94 (1.73 – 2.37) and 0.88 – 0.94 (1.83 – 2.67) respectively. At mCAFT, the Zephyr ICCs (SEM) ranged from 0.91 – 0.97 (3.12 – 4.64) and 0.85 – 0.98 (3.28 – 4.88) for the Fitbit. Throughout the recovery, the ICCs (SEM) ranged from 0.93 – 0.97 (2.65 – 4.66) and 0.76 – 0.91 (3.17 – 4.67) for Zephyr and Fitbit devices respectively. Pearson’s correlation coefficients and (Mean differences) for heart rate variable were 0.97 – 0.99 (-0.60 – 0.02) at Rest, 0.89 – 0.99 (13.51 – 0.62) at submaximal testing and 0.70 – 0.84 (-0.54 – 2.52) throughout recovery. The average agreement bias of heart rate in pair-wise device comparison indicated mean differences of -0.20, 4.00 and 1.00 at rest, sub-maximal testing and recovery respectively. We identified fair to very good quality evidence from 14 studies. The Zephyr Bioharness and Fitbit Charge devices demonstrated excellent reliability measures, and the Fitbit Charge device heart rate variable demonstrated strong to very strong correlations when concurrently compared with Zephyr, and provided valuable information regarding its interchangeable use in a sample of sixty healthy male and female participants of various age groups during a resting, standardized submaximal fitness and recovery phases. / Thesis / Master of Science (MSc)