No description available.

Continuous scaling of the transistor size and reduction of the operating voltage have led to a significant performance improvement of integrated circuits. However, the vulnerability of the scaled circuits to transient data upsets or soft errors, which are caused by alpha particles and cosmic neutrons, has emerged as a major reliability concern. In this thesis, we have investigated the effects of soft errors in combinational circuits and proposed soft error detection techniques for high speed adders. In particular, we have proposed an area-efficient 64-bit soft error robust logarithmic adder (SRA). The adder employs the carry merge Sklansky adder architecture in which carries are generated every 4 bits. Since the particle-induced transient, which is often referred to as a single event transient (SET) typically lasts for 100~200 ps, the adder uses time redundancy by sampling the sum outputs twice. The sampling instances have been set at 110 ps apart. In contrast to the traditional time redundancy, which requires two clock cycles to generate a given output, the SRA generates an output in a single clock cycle. The sampled sum outputs are compared using a 64-bit XOR tree to detect any possible error. An energy efficient 4-input transmission gate based XOR logic is implemented to reduce the delay and the power in this case. The pseudo-static logic (PSL), which has the ability to recover from a particle induced transient, is used in the adder implementation. In comparison with the space redundant approach which requires hardware duplication for error detection, the SRA is 50% more area efficient. The proposed SRA is simulated for different operands with errors inserted at different nodes at the inputs, the carry merge tree, and the sum generation circuit. The simulation vectors are carefully chosen such that the SET is not masked by error masking mechanisms, which are inherently present in combinational circuits. Simulation results show that the proposed SRA is capable of detecting 77% of the errors. The undetected errors primarily result when the SET causes an even number of errors and when errors occur outside the sampling window.

Tests for detecting faults in analog and mixed-signal circuits have been traditionally derived from the datasheet speci and #64257;cations. Although these speci and #64257;cations describe important aspects of the device, in many cases these application oriented tests are costly to implement and are inefficient in determining product quality. Increasingly, the gap between speci and #64257;cation test requirements and the capabilities of test equipment has been widening. In this work, a systematic method to generate and evaluate alternate tests for detecting parametric faults is proposed. We recognize that certain aspects of analog test generation problem are not amenable to automation. Additionally, functional features of analog circuits are widely varied and cannot be assumed by the test generator. To overcome these problems, an extended device under test (DUT) model is developed that encapsulates the DUT and the DUT speci and #64257;c tasks. The interface of this model provides a well de and #64257;ned and uniform view of a large class of devices. This permits several simpli and #64257;cations in the test generator. The test generator is uses a search-based procedure that requires evaluation of a large number of candidate tests. Test evaluation is expensive because of complex fault models and slow fault simulation techniques. A tester-resident test evaluation technique is developed to address this issue. This method is not limited by simulation complexity nor does it require an explicit fault model. Making use of these two developments, an efficient and automated test generation method is developed. Theoretical development and a number of examples are used to illustrate various concepts that are presented in this thesis.

This thesis considers the analysis of system performance and resource allocation for a coded OFDM system over frequency selective fading channels. Due to the inseparable role taken by channel coding in a coded OFDM system, an information theoretical analysis is carried out and taken as the basis for the system performance and throughput. Based on the results of the information theoretical analysis, the optimal system BER performance of a coded OFDM system is first shown to converge to the outage probability for large OFDM block lengths. Instead of evaluating the outage probability numerically, we provide in this thesis a simple analytical closed form approximation of the outage probability for a coded OFDM system over frequency selective quasi-static fading channels. Simulation results of the turbo-coded OFDM systems further confirm the approximation of the outage probability. By taking the instantaneous channel capacity as the analytical building block, system throughput of a coded OFDM system is then provided. With the aim to compare the performance difference between adaptive and uniform resource allocation strategies, the system throughput of different allocation schemes under various channel conditions is analyzed. First, it is demonstrated that adaptive power allocation over OFDM sub-carriers at the transmitter achieves very little gain in terms of throughput over a uniform power distribution scheme. Theoretical analysis is then provided of the throughput increase of adaptive-rate schemes compared with fixed-rate schemes under various situations. Two practical OFDM systems implementing rate-compatible-punctured-turbo-code-based (RCPT-based) hybrid automatic-repeat-request (Hybrid-ARQ) and redundancy incremental Hybrid-ARQ protocols are also provided to verify the analytical results.

One of the most basic reaction time experiments employed by psychologists is the comparison of latencies to responses for single and redundant targets. The general effect is that participants are capable of responding faster, that is having shorter response latencies when redundant stimuli, as opposed to an individual stimulus, are presented. Interestingly, several models attempting to predict this effect, including the well known race model, have not been entirely successful. The following study evaluated redundancy gain and violations of the race model, in three experimental models: visual only, auditory only, and a visual-auditory bimodal paradigm. The results showed redundancy gain in all three paradigms, but they were only significant violations of the race model for the visual-auditory condition. Additionally, correlations between the different paradigms were explored with respect to redundancy gain and violations of the race model on an individual participant basis.

This work studies application acceleration for wireless and mobile data networks. The problem of accelerating application can be addressed along multiple dimensions. The first dimension is advanced network protocol design, i.e., optimizing underlying network protocols, particulary transport layer protocol and link layer protocol. Despite advanced network protocol design, in this work we observe that certain application behaviors can fundamentally limit the performance achievable when operating over wireless and mobile data networks. The performance difference is caused by the complex application behaviors of these non-FTP applications. Explicitly dealing with application behaviors can improve application performance for new environments. Along this overcoming application behavior dimension, we accelerate applications by studying specific types of applications including Client-server, Peer-to-peer and Location-based applications. In exploring along this dimension, we identify a set of application behaviors that significantly affect application performance. To accommodate these application behaviors, we firstly extract general design principles that can apply to any applications whenever possible. These design principles can also be integrated into new application designs. We also consider specific applications by applying these design principles and build prototypes to demonstrate the effectiveness of the solutions. In the context of application acceleration, even though all the challenges belong to the two aforementioned dimensions of advanced network protocol design and overcoming application behavior are addressed, application performance can still be limited by the underlying network capability, particularly physical bandwidth. In this work, we study the possibility of speeding up data delivery by eliminating traffic redundancy present in application traffics. Specifically, we first study the traffic redundancy along multiple dimensions using traces obtained from multiple real wireless network deployments. Based on the insights obtained from the analysis, we propose Wireless Memory (WM), a two-ended AP-client solution to effectively exploit traffic redundancy in wireless and mobile environments. Application acceleration can be achieved along two other dimensions: network provision ing and quality of service (QoS). Network provisioning allocates network resources such as physical bandwidth or wireless spectrum, while QoS provides different priority to different applications, users, or data flows. These two dimensions have their respective limitations in the context of application acceleration. In this work, we focus on the two dimensions of overcoming application behavior and Eliminating traffic redundancy to improve application performance. The contribution of this work is as follows. First, we study the problem of application acceleration for wireless and mobile data networks, and we characterize the dimensions along which to address the problem. Second, we identify that application behaviors can significantly affect application performance, and we propose a set of design principles to deal with the behaviors. We also build prototypes to conduct system research. Third, we consider traffic redundancy elimination and propose a wireless memory approach.

<p>The increasing number of biological databases today requires that users are able to search more efficiently among as well as in individual databases. One of the most widespread problems is redundancy, i.e. the problem of duplicated information in sets of data. This thesis aims at implementing an algorithm that distinguishes from other related attempts by using the genomic positions of sequences, instead of similarity based sequence comparisons, when making a sequence data set non-redundant. In an automatic updating procedure the algorithm drastically increases the possibility to update and to maintain the topicality of a non-redundant database. The procedure creates a biologically sound non-redundant data set with accuracy comparable to other algorithms focusing on making data sets non-redundant</p>

Modern day bridge foundations are often designed for the maximum loading condition on a single member and each following foundation member often follows the same design. A 4-pile and a 2-pile foundation system were analyzed with an upperbound plasticity model in an attempt to approximately represent the system capacity. A sensitivity analysis was performed to measure the change in the ultimate system capacity by altering the parameters used to define individual pile capacities. The load and the resistance are considered to be normally distributed and are each characterized by a mean and a standard deviation. A first order reliability method was used to measure and account for any system redundancy in the foundation system by predicting a probability of failure based on the load and resistance of the system. Based on the most heavily loaded pile and the analysis performed on the 2-pile and 4-pile foundation systems, the probabilities of failure are lower for the system than for the single pile. This would bring reason to state there is some measurable redundancy in the basic systems analyzed where these systems would once be considered to be non-redundant systems. / text

Motor noise and redundancy are vexing issues in motor control; yet their understanding provides great insights on underlying control mechanisms that govern movement. They provide glimpses into how the nervous system organizes and regulates movement within the motor control system. Understand of motor control could spur new advances in motor control could lead to better development of rehabilitation process and technology to counteract debilitating affects of neuromuscular disorders and motor readjustment with prostheses. However, before such process and technology could be developed and adapted for clinical use, a deeper understanding of motor control is needed to unravel the neural roadmap that regulates and generates movement. New theory of motor control could precipitate the development of more robust control mechanisms for robotic-human interaction. This work aims at expanding a more rigorous analytical and mathematical framework to understand how these control mechanisms reconcile redundancy and stochastic noise in human motor control. / text

Trial-to-trial variability in human movement is often overlooked and averaged out, but useful information can be gleaned on the brain’s control of variability. A task can be defined by a function specifying a solution manifold along which all task variable combinations will lead to goal success – the Goal-Equivalent Manifold (GEM). We selected a reaching task with variables reach Distance (D) and reach Time (T). Two GEMs were selected: a constant D/T and constant D×T. Subjects had no knowledge of the goal prior to the experiments and were instructed only to minimize error. Subjects learned the generalized tasks by reducing errors and consolidated learning from one day to the next, generalized learning from the D×T to the D/T GEM, and had interference of learning from the D/T to the D×T GEM. Variability was structured along each GEM significantly more than perpendicular to it. Deviations resulting in errors were corrected significantly more quickly than any other deviation. Our results indicate that subjects can learn generalized reaching tasks, and the brain exploits redundancy in those tasks. / text