Modern telecommunications companies are moving away from conventional circuit-switched architectures to more versatile packet-switched infrastructures. Traditional First-In-FirstOut (FIFO) queues that are currently multiplexing IP traffic are not able to meet the strict Quality-of-Service (QoS) requirements of delay sensitive real-time traffic. Two main solution families exist that separate heterogeneous traffic into appropriate classes. The first is known as Generalized Processor Sharing (GPS), which divides the available bandwidth among the contending classes, proportionally to the throughput guarantee negotiated with each class. GPS and its myriad of packetised variants are relatively easy to analyse, as the service rate of individual classes is directly related to its throughput guarantee. As GPS splits the arriving traffic into separate queues, it is useful for best-effort traffic, supplying each class of traffic with either a maximum or minimum amount of bandwidth that it deserves. The second solution is the Earliest Deadline First (EDF) scheduler, also known as Earliest Due Date (EDD). Each traffic class has a delay deadline, by which the individual packets need to be served in order to meet their heterogeneous QoS requirements. EDF selects packets that are closest to their deadline. It is therefore primarily useful for delay sensitive real-time traffic. Although this is a simple algorithm, it turns out to be surprisingly difficult to analyse. Several papers attempted to analyse EDF. Most of them found either discrete bounds, which lie far away from the mean, or stochastic bounds which tend to capture the delay behaviour of the traffic more accurately. After the introductory first chapter, this thesis simulates a realistic cellular environment, where packets of various classes of service are transmitted across an HSDPA air interface. The aim is to understand the behaviour of EDF and its channel aware Opportunistic EDF scheduler compared to other scheduling families commonly used in HSDPA environments. In particular, Round Robin is simulated as the most simplistic scheduler. Max ell chooses packets solely based on the best channel conditions. Finally, PF -T is a scheme that tries to maximise the overall transmission rate that packets experience, but this metric gets divided by the throughput that each class already achieved. This introduces a form of long-term fairness that prevents the starvation of individual classes. The third chapter contains the main analysis, which uses Large Deviation principles and the Effective Bandwidth theory to approximate the deadline violation probability and the delay density function of EDF in a wired network. A definition for the fairness of EDF is proposed. The analysis is extended to approximate the stochastic fairness distribution. In the fourth chapter of the thesis an opportunistic EDF scheduler is proposed for mobile legs of a network that takes advantage of temporary improvements in the channel conditions. An analytical model is developed that predicts the delay density function of the opportunistic EDF scheduler. The channel propagation gain is assumed to be log-normally distributed, which requires graphical curve fitting, as no closed-form solution exists / Thesis (Ph.D.)-University of KwaZulu-Natal, Durban, 2009.

This dissertation focuses on the application of tomography in the sugar milling process, specifically within the vacuum pan. The research aims to improve the efficiency and throughput of a sugar mill by producing real-time images of the boiling dynamic in the pan and hence can be used as a diagnostic tool. The real-time tomography system is a combination of ruggedized data collecting hardware, a switching circuit and software algorithms. The system described in this dissertation uses 16 electrodes and estimates images based on the distinct differences in conductivities to be found in the vacuum pan, i.e. a conductive syrup-like fluid (massecuite) and bubbles. There is a direct correlation between the bubbles produced during the boiling process and heat transfer in the pan. From this correlation one can determine how well the pan is operating. The system has been developed in order to monitor specific parts of a pan for optimal boiling. A binary reconstructed image identifies either massecuite or water vapour. Each image is reconstructed using a modified neighbourhood data collection method and a back projection algorithm. The data collection and image reconstruction take place simultaneously, making it possible to generate images in real-time. Each image frame is reconstructed at approximately 1.1 frames per second. Most of the system was developed in LabVIEW, with some added external drive electronics, and functions seamlessly. The tomography system is LAN enabled hence measurements are initiated through a remote PC on the same network and the reconstructed images are streamed to the user. The laboratory results demonstrate that it is possible to generate tomographic images from bubbles vs massecuite, tap water and deionized water in real-time. / Thesis (M.Sc.Eng.)-University of KwaZulu-Natal, Durban, 2012.

No description available.

A real-time system is a system such that the correctness of operations depends not only on the logical results, but also on the time at which these results are available. A fundamental problem in designing real-time systems is to analyze response time of operations, which is defined as the time elapsed from the moment when the operation is requested to the moment when the operation is completed. Response time analysis is challenging due to the complex dynamics among operations. A common technique is to study response time under worst-case scenario. However, using worst-case response time may lead to the conservative real-time system designs. To improve the real-time system design, we analyze the non-worst-case response time of operations and apply these results in the design process. The main contribution of this thesis includes mathematical modeling of real-time systems, calculation of non-worst-case response time, and improved real-time system design. We perform analysis and design on three common types of real-time systems as the real-time computing system, real-time communication network, and real-time energy management. For the real-time computing systems, our non-worst-response time analysis leads a necessary and sufficient online schedulability test and a measure of robustness of real-time systems. For the real-time communication network, our non-worst-response time analysis improves the performance for the model predictive control design based on the real-time communication network. For the real-time energy management, we use the non-worst-case response time to check whether the micro-grid can operate independently from the main grid.

Real-time systems are amongst the most safety critical systems involving computer software and the incorrect functioning of this software can cause great damage, up to and including the loss of life. If seems sensible therefore to write real-time software in a way that gives us the best chance of correctly implementing specifications. Because of the high level of functional programming languages, their semantic simplicity and their amenability to formal reasoning and correctness preserving transformation it thus seems natural to use a functional language for this task. This thesis explores the problems of applying functional programming languages to real-time by defining the real-time functional programming language Ruth. The first part of the thesis concerns the identification of the particular problems associated with programming real-time systems. These can broadly be stated as a requirement that a real-time language must be able to express facts about time, a feature we have called time expressibility. The next stage is to provide time expressibility within a purely functional framework. This is accomplished by the use of timestamps on inputs and outputs and by providing a real-time clock as an input to Ruth programs. The final major part of the work is the construction of a formal definition of the semantics of Ruth to serve as a basis for formal reasoning and transformation. The framework within which the formal semantics of a real-time language are defined requires time expressibility in the same way as the real-time language itself. This is accomplished within the framework of domain theory by the use of specialised domains for timestamped objects, called herring-bone domains. These domains could be used as the basis for the definition of the semantics of any real-time language.

Thesis (Ph. D.)--University of Alabama at Birmingham, 2007. / Additional advisors: Jeff G. Gray, Marjan Mernik, Rajeev Raje, Chengcui Zhang. Description based on contents viewed Feb. 7, 2008; title from title screen. Includes bibliographical references (p. 216-230).

Thesis (M.S.)--University of Missouri--Rolla, 2007. / Vita. The entire thesis text is included in file. Title from title screen of thesis/dissertation PDF file (viewed February 14, 2008) Includes bibliographical references (p. 37-40).

Thesis (M.S.)--Rochester Institute of Technology, 1989. / Includes bibliographical references (leaf 85).

Thesis (Ph.D.) -- University of Western Sydney, 2006. / A thesis presented to the School of Computing and Mathematics, University of Western Sydney, in fulfilment of the requirements for the degree of Doctor of Philosophy. Includes bibliography.

Thesis (Ph.D.)--University of Wollongong, 2004. / Typescript. Includes bibliographical references: leaf 151-159.