Thesis (MScIng)--University of Stellenbosch, 2005. / Bibliography. Also available via the Internet.

Thesis (Ph.D.)--University of Delaware, 2007. / Principal faculty advisor: Charles G. Boncelet, Dept. of Computer & Information Sciences. Includes bibliographical references.

This thesis is concerned with the adaptive adjustment of digital data receivers employed m synchronous serial data transmission systems that use quadrature amplitude modulation The receiver employs a pre-detection filter that forms the first part of a decision feedback equalizer or else is used ahead of a near maximum-likelihood detector. The filter is ideally adjusted such that the sampled impulse response of the channel and filter is minimum phase.

A monochromatic image can be represented by a real function f(x,y,t) of two spatial variables and time. Picture transmission research is directed at finding methods of transmitting a facsimile of f(x,y,t) at minimum cost which still satisfies some quality criteria. The ensemble of source, functions, and the measures of cost and quality are defined by the application. For many applications, the subjective quality is an important measure of performance. Generally, it is difficult if not impossible to find a tractable analytic relationship between the source and system parameters, and the subjective quality. Thus,for picture processing research, equipment must be available for simulating proposed systems so that pictures can be produced and estimates made of the subjective quality. This thesis will describe a system which was devised for studying low resolution picture transmission systems. The system employs an image dissector to input picture data to a PDP-9 computer. The data can be stored on magnetic tape, processed and then displayed on a picture output device which employs a Tektronix 561 Oscilloscope. A set of general purpose programs was developed for manipulating picture data using this hardware configuration. Three test pictures were read in and particular programs were written to calculate some picture statistics. The statistics include signal histograms, auto-covariance functions and, some run length statistics for the two level material. These statistics were then used as design information for some simulation programs. A program was written to simulate Pulse Code Modulation (PCM) and Differential Pulse Code Modulation (DPCM) systems in order to evaluate the subjective effect of channel errors on pictures transmitted by PCM and DPCM. An attempt was also made to develop a simple two-dimensional filter program to reduce the effect of channel noise. Since the channel noise in the PCM and DPCM systems produced two-dimensional noises which were very different in appearance, two filtering routines were necessary. Subjective rating tests were conducted to evaluate the filtered and unfiltered PCM and DPCM pictures. For the unfiltered systems, these tests indicate that for good channels, the DPCM system is better than the PCM system while for very bad channels, the pictures produced by the PCM system are subjectively better than those of a DPCM system using the same number of bits. It was found also that the PCM filter proposed, improved the noisy pictures while the proposed DPCM filter produced no improvement and even lowers the quality of some of the pictures. / Applied Science, Faculty of / Electrical and Computer Engineering, Department of / Graduate

We present a design for a remote database access facility, which uses a message system as its communication medium. Adopting a message-based design offers a number of advantages over more conventional connection-oriented architectures for those remote database applications where the interaction between user and DBMS involves a single query followed by a single result. By complying with the CCITT X.400 Recommendations on Message Handling Systems, the design allows for networkwide access to remote DBMSs that is independent of the nature of DBMSs being accessed, the systems on which they reside and the network over which they are accessed. An initial implementation using the EAN distributed message system, developed at the University of British Columbia, is described as a means of demonstrating the design's feasibility. / Science, Faculty of / Computer Science, Department of / Graduate

Approved for public release; distribution is unlimited / A study of SONET network management applications and the load they impart to the network is conducted to provide a better understanding of the capability of various management approaches. In this study, a SONET network is set up in the Advanced Networking Laboratory of the Naval Postgraduate School using four Cisco ONS 15454s. Next, two Element Management Systems, the Cisco Transport Controller and the Cisco Transport Manager, are deployed onto the SONET network. Subsequently, the network traffic of the Element Management Systems is captured and analyzed using a packet analyzer. Link utilization of the two tools is computed using the first-order statistics of the captured traffic distributions. In addition, the Hurst parameter is estimated using the variance-index plot technique (which uses higher-orders statistics of the modeled distributions) to determine the captured traffic's degree of self-similarity. Finally, the calculated utilization is extrapolated to obtain the link utilization for 2500 network elements (the maximum number supported by the Cisco Transport Manager). The result obtained is useful in determining the maximum number of network elements (Cisco ONS 15454s) that the Cisco Transport Manager can support from a network loading point of view. / Civilian, Ministry of Defense, Singapore

An evaluation of the performance of a SONET management system was conducted to better understand its management capabilities due to network disruptions in the presence of a traffic load. This study analyzed the Cisco Transport Manager (CTM) which manages a testbed of four Cisco ONS15454 optical systems. The network was injected with HTTP and FTP traffic generated by the Spirent Smartbits system installed with TeraMetrics Gigabit Ethernet modules and load calibration configured by the Spirent Avalanche software. To simulate real-world situations, power disruptions were applied to the network while collecting CTM traffic using Ethereal. Using queuing analysis, the arrival rates and service times were computed for various CTM traffic components and a utilization for 2500 network elements (NE) extrapolated. Self-similarity analysis was performed and the logvariance was plotted to extract the Hurst values. Finally, the results and findings were compared with prior research for loading and no-loading cases. The results of this study are useful in determining the maximum number of network elements manageable in a disruptive environment. Final analysis on the effects of link utilization on the queue size showed that the CTM is able to manage more NEs when the network is disrupted. Unfortunately, managing more NEs increases the queue size even though the utilization was found to be 0.83 for 5450 NEs. Consequently, in order to maintain a moderate queue size, the maximum number of NEs manageable was found to be 2495. This value is close to CISCO's specification of a CTM server managing a maximum of 2500 NEs.

Thesis (Ph. D.)--University of Washington, 2003. / Vita. Includes bibliographical references (leaves 83-88).

No description available.

Thesis (M.S.)--Rochester Institute of Technology, (1989). / Includes bibliographical references (leaves 170-171).