Browsing by Author "Dr. Alexander Dean, Committee Member"

Now showing 1 - 8 of 8

Compositional Static Cache Analysis Using Module-level Abstraction
(2003-12-10) Patil, Kaustubh Sambhaji; Dr. Frank Mueller, Committee Chair; Dr. Alexander Dean, Committee Member; Dr. Eric Rotenberg, Committee Member
Static cache analysis is utilized for timing analysis to derive worst-case execution time of a program. Such analysis is constrained by the requirement of an inter-procedural analysis for the entire program. But the complexity of cycle-level simulations for entire programs currently restricts the feasibility of static cache analysis to small programs. Computationally complex inter-procedural analysis is needed to determine caching effects, which depend on knowledge of data and instruction references. Static cache simulation traditionally relies on absolute address information of instruction and data elements. This thesis presents a framework to perform worst-case static cache analysis for direct-mapped instruction caches using a module-level and compositional approach, thus addressing the issue of complexity of inter-procedural analysis for an entire program. The module-level analysis parameterizes the data-flow information in terms of the starting offset of a module. The compositional analysis stage uses this parameterized data-flow information for each module. Thus, the emphasis here is on handling most of the complexity in the module-level analysis and performing as little analysis as possible at the compositional level. The experimental results show that the compositional analysis framework provides equally accurate predictions when compared with the simulation approach that uses complete inter-procedural analysis.
A Conformance Review Strategy for Regulating Safety-Critical Software.
(2006-09-24) Jetley, Raoul Praful; Dr. S. Purushothaman Iyer, Committee Chair; Dr. Laurie Williams, Committee Member; Dr. Matthias Stallmann, Committee Member; Dr. Alexander Dean, Committee Member
Safety is an important concern for software used in life-critical systems such as air transport, nuclear power and medical devices. The critical nature of these systems necessitates that the software used therein be reliable and free of errors. It becomes imperative, therefore, to have a stringent review process in place to ascertain the correctness of the software and to ensure that it meets all requirements and standards. Regulatory agencies encourage the use of formal methods based techniques in the development of safety critical software. However, most manufacturers are reluctant to use these techniques, citing them as too complex and time consuming. As a result, (potentially life-threatening) errors are often not discovered until the software is already on the market. When such an error is eventually discovered, it becomes essential to trace the failure to its exact source in the implementation and to assure that the error correction restores the overall safety and effectiveness of the device. In this dissertation, we present how efficient premarket and postmarket reviews of designs and implementations can be carried out using formal methods based techniques, to enable the process of reviewing software in safety-critical devices. To facilitate premarket conformance reviews, we introduce the notion of usage models -- standardized formal models that serve as design templates. We present an approach to conformance checking of safety-critical software through formal verification and automated test case sequences derived from these standardized models. To provide for efficient postmarket reviews, we establish a methodology based on integrating program slicing with model abstraction to trace software failures to their root cause. We formalize this methodology by presenting an iterative algorithm for abstraction-driven slicing and realize this algorithm through the implementation of the CAdS -- a forensic analysis tool for C programs. We provide case studies involving typical medical device software to illustrate the various concepts involved and present results from these studies to gauge the effectiveness of our proposed approach.
Design of a Flexible DSP Based Controller Hardware System for Power Electronics Applications
(2008-07-27) Godbole, Rahul Pushpak; Dr. Alexander Dean, Committee Member; Dr. Alex Huang, Committee Member; Dr. Subhashish Bhattacharya, Committee Chair
The Effects of Gain Adaptation for QoS Deterioration in Internet-based Teleoperation Involving Use of a Virtual Reality Interface
(2003-04-22) Sheik Nainar, Mohamed Ashraf Ali; Dr. David Kaber, Committee Co-Chair; Dr. Mo-yuen Chow, Committee Co-Chair; Dr. Alexander Dean, Committee Member
The goals of this study included evaluating the affects of different types of communication network delays on remote-control rover (telerover) performance, operator telepresence experiences and workload. The study also evaluated the utility of gain adaptation for communication delays on telepresence, performance and workload. Finally, the work examined the relationship between performance and subjective presence in an Internet-based teleoperation scenario utilizing a virtual reality (VR) interface. Telepresence has been identified as a design ideal for teleoperation systems, and task environment factors, such as disturbances in human-machine interaction, have been identified as potential underpinnings of presence experiences. A VR-based simulation of a telerover navigation task was developed for this study. The task involved navigating the rover in a virtual environment (VE) between obstacles, like a slalom ski race. Task performance measures included time-to-task completion (TTC) and the number of collisions of the rover with task obstacles (errors). Two levels of telerover automation (LOA) were implemented including teleoperation, or manual control, and telerobotic or automation assisted control. Combinations of LOAs and delay types, including constant and random, were tested with and without gain adaptation. A mixed between-within experimental design was used in which LOA served as a grouping variable. Each subject experienced 10 test trials (2 no-delay + 2 ( (2 delay types ( 2 adaptation settings)) under either the teleoperation or telerobotic control mode. Subject exposure to various network conditions was randomized. Presence questionnaire and the NASA Task Load Index were used to capture subjective telepresence and workload ratings, respectively, at the end of each test trial. Results revealed that LOA, delay, adaptation and the interaction of LOA and a variable describing the overall network condition (a combination of delay and adaptation) significantly effected TTC. The telerobotic control mode produced the best TTC irrespective of the delay type and adaptation. Both delay types combined with adaptation produced the worst TTC within each LOA, as compared to all other network conditions. Performance errors/collisions were significantly effected by LOA, delay and adaptation. The telerobotic control mode produced the greatest number of errors and the adaptation conditions were superior to no-adaptation conditions. The constant delay produced more errors than the random delay type. Both telepresence and workload were significantly affected by LOA and individual differences with telerobotic control producing higher telepresence ratings along with lower workload scores. Telepresence was found to be significantly correlated with TTC, specifically there was a reduction in TTC with an increase in telepresence ratings. Workload was significantly positively correlated with telepresence. Although the telerobotic control mode reduced operator workload, it off-loaded some of the rover control responsibilities from the user to the machine system allowing the operator to pay more attention to the VR displays promoting their knowledge of the current state of the VE and possibly presence sensations. These correlation analysis results are similar to those established by previous research. It was expected that the gain adaptation would better support users in achieving and sustaining telepresence. Although changes in telepresence across the adaptation and no-adaptation conditions under telerobotic control were inline with this hypothesis, similar results were not found with teleoperation control. The results of this study are directly applicable to the selection of guaranteed communication network parameters through Quality of Service (QoS) in Internet-based telemanipulation systems. It also can be used as a guideline for telerover control mode selection for time and error critical teleoperation. Finally, the results support the notion that telepresence may be important to performance in teleoperation tasks (and that gain adaptation for network delays under certain control modes may be beneficial to telepresence).
Frequency-aware Static Timing Analysis for Power-aware Embedded Architectures
(2004-03-14) Seth, Kiran Ravi; Dr. Frank Mueller, Committee Chair; Dr. Alexander Dean, Committee Member; Dr. Eric Rotenberg, Committee Member
Power is a valuable resource in embedded systems as the lifetime of many such systems is constrained by their battery capacity. Recent advances in processor design have added support for dynamic frequency/voltage scaling (DVS) for saving power. Recent work on real-time scheduling focuses on saving power in static as well as dynamic scheduling environments by exploiting idle and slack due to early task completion for DVS of subsequent tasks. These scheduling algorithms rely on a priori knowledge of worst-case execution times (WCET) for each task. They assume that DVS has no effect on the worst-case execution cycles (WCEC) of a task and scale the WCET according to the processor frequency. However, for systems with memory hierarchies, the WCEC typically does change under DVS due to frequency modulation. Hence, current assumptions used by DVS schemes result in a highly exaggerated WCET. The research presented contributes novel techniques for tight and flexible static timing analysis particularly well-suited for dynamic scheduling schemes. The technical contributions are as follows: (1) The problem of changing execution cycles due to scaling techniques is assessed. (2) A parametric approach towards bounding the WCET statically with respect to the frequency is proposed. Using a parametric model, the effect of changes in frequency on the WCEC can be captured and, thus, the WCET over any frequency range can be accurately modeled. (3) The design and implementation of the frequency-aware static timing analysis (FAST) tool, based on prior experience with static timing analysis, is discussed. (4) Experiments demonstrate that the FAST tool provides safe upper bounds on the WCET, which are tight. The FAST tool allows the capture of the WCET of six benchmarks using equations that overestimate the WCET by less than 1%. FAST equations can also be used to improve existing DVS scheduling schemes to ensure that the effect of frequency scaling on WCET is considered and that the WCET used is not exaggerated. (5) Three DVS scheduling schemes are leveraged by incorporating FAST into them and by showing that the power consumption further decreases.
A New Operating System and Application Programming Interface for the EvBot Robot Platform
(2010-04-27) Colon, Micah; Dr. Edward Grant, Committee Chair; Dr. Alexander Dean, Committee Member; Dr. Troy Nagle, Committee Member
The research presented in this thesis describes the development of the Linux distribution and a new control architecture for robots. The reasons Linux was chosen are enumerated and a description of the build system and setup used to generate the distribution, with support for multiple platforms, is discussed. The Evbot Abstraction Layer (EAL), a new robot control architecture and framework is described, and the simple API is detailed.
A Pneumatically Actuated Brace Designed For Upper Extremity Stroke Rehabilitation
(2003-07-07) Merritt, Carey Reid; Dr. Mark White, Committee Member; Dr. Edward Grant, Committee Chair; Dr. Alexander Dean, Committee Member
Nearly 700,000 people suffered from stroke last year and those who survived were left with any number of disabilities. One of the most common disabilities in stroke is paralysis of the upper arm. Since therapy for this disability is expensive, patients are finding rehabilitation difficult to afford and manage. This thesis proposes an inexpensive pneumatic wearable garment for the patient to use for stroke rehabilitation. Unlike most rehabilitation robots, which are large, non-compliant, and expensive, this device will enable the patient to purchase the garment and move freely within their own home while rehabilitating their affected arm. In this thesis, a wearable elbow device similar to the proposed wearable garment was designed using an inexpensive elbow brace. The elbow brace used custom made artificial air muscles also known as McKibben Artificial Muscles to substitute for the biceps and triceps, which are responsible for flexion and extension of the human elbow. These artificial muscles were chosen for their low-cost, compliance, lightweight, and large force capabilities. The air muscles were designed and developed especially for this device and cost less than $3.00 to make and weigh approximately 11 g. This pneumatically actuated elbow brace was controlled using solenoid valves in conjunction with a Mitsubishi M32/83C 16-bit micro controller to achieve flexion and extension of the elbow. Experiments on the pneumatic elbow brace have shown that it is capable of moving a passive patient's arm within a 110° range, which is adequate for rehabilitation of the elbow.
Weld for Itanium Processor
(2002-12-03) Sharma, Saurabh; Dr. Thomas M. Conte, Committee Chair; Dr. Eric Rotenberg, Committee Member; Dr. Alexander Dean, Committee Member
This dissertation extends a WELD for Itanium processors. Emre Özer presented WELD architecture in his Ph.D. thesis. WELD integrates multithreading support into an Itanium processor to hide run-time latency effects that cannot be determined by the compiler. Also, it proposes a hardware technique called operation welding that merges operations from different threads to utilize the hardware resources. Hardware contexts such as program counters and the fetch units are duplicated to support for multithreading. The experimental results show that Dual-thread WELD attains a maximum of 11% speedup as compared to single-threaded Itanium architecture while still maintaining the hardware simplicity of the EPIC architecture.