Delivering high-quality of video to end users over the best-effort Internet is a challenging task since quality of streaming video is highly subject to network conditions. A fundamental issue in this area is how real-time applications cope with network dynamics and adapt their operational behavior to offer a favorable streaming environment to end users. As an effort towards providing such streaming environment, the first half of this work focuses on analyzing the performance of video streaming in best-effort networks and developing a new streaming framework that effectively utilizes unequal importance of video packets in rate control and achieves a near-optimal performance for a given network packet loss rate. In addition, we study error concealment methods such as FEC (Forward-Error Correction) that is often used to protect multimedia data over lossy network channels. We investigate the impact of FEC on the quality of video and develop models that can provide insights into understanding how inclusion of FEC affects streaming performance and its optimality and resilience characteristics under dynamically changing network conditions. In the second part of this thesis, we focus on measuring bandwidth of network paths, which plays an important role in characterizing Internet paths and can benefit many applications including multimedia streaming. We conduct a stochastic analysis of an end-to-end path and develop novel bandwidth sampling techniques that can produce asymptotically accurate capacity and available bandwidth of the path under non-trivial cross-traffic conditions. In addition, we conduct comparative performance study of existing bandwidth estimation tools in non-simulated networks where various timing irregularities affect delay measurements. We find that when high-precision packet timing is not available due to hardware interrupt moderation, the majority of existing algorithms are not robust to measure end-to-end paths with high accuracy. We overcome this problem by using signal de-noising techniques in bandwidth measurement. We also develop a new measurement tool called PRC-MT based on theoretical models that simultaneously measures the capacity and available bandwidth of the tight link with asymptotic accuracy.

"As streaming techniques and wireless access networks become more widely deployed, a streaming multimedia connection with the "last mile" being a wireless network is becoming increasingly common. However, since current streaming techniques are primarily designed for wired networks, streaming multimedia applications can perform poorly in wireless networks. Recent research has shown that the wireless network conditions, such as the wireless link layer rate adaptation, contending traffic, and interference can significantly degrade the performance of streaming media applications. This performance degradation includes increased multimedia frame losses and lower image quality caused by packet loss, and multiple rebuffering events that stop the media playout. This dissertation presents the model, design, implementation and evaluation of an application layer solution for improving streaming multimedia application performance in IEEE 802.11 wireless networks by using enhanced bandwidth estimation techniques. The solution includes two parts: 1) a new Wireless Bandwidth estimation tool (WBest) designed for fast, non-intrusive, accurate estimation of available bandwidth in IEEE 802.11 networks, which can be used by streaming multimedia applications to improve the performance in wireless networks; 2) a Buffer and Rate Optimization for Streaming (BROS) algorithm using WBest to guide the streaming rate selection and initial buffer optimization. WBest and BROS are implemented and incorporated into an emulated streaming client-server system, Emulated Streaming (EmuS), in Linux and evaluated under a variety of wireless conditions. The evaluations show that with WBest and BROS, the performance of streaming multimedia applications in wireless networks can be significantly improved in terms of multimedia frame loss, rebuffer events and buffer delay."

No description available.

No description available.

With the advent of the internet and advances in computing power, the collection of very large high-dimensional datasets has become feasible { understanding and modelling high-dimensional data has thus become a crucial activity, especially in the field of pattern recognition. Since non-parametric density estimators are data-driven and do not require or impose a pre-defined probability density function on data, they are very powerful tools for probabilistic data modelling and analysis. Conventional non-parametric density estimation methods, however, originated from the field of statistics and were not originally intended to perform density estimation in high-dimensional features spaces { as is often encountered in real-world pattern recognition tasks. Therefore we address the fundamental problem of non-parametric density estimation in high-dimensional feature spaces in this study. Recent advances in maximum-likelihood (ML) kernel density estimation have shown that kernel density estimators hold much promise for estimating nonparametric probability density functions in high-dimensional feature spaces. We therefore derive two new iterative kernel bandwidth estimators from the maximum-likelihood (ML) leave one-out objective function and also introduce a new non-iterative kernel bandwidth estimator (based on the theoretical bounds of the ML bandwidths) for the purpose of bandwidth initialisation. We name the iterative kernel bandwidth estimators the minimum leave-one-out entropy (MLE) and global MLE estimators, and name the non-iterative kernel bandwidth estimator the MLE rule-of-thumb estimator. We compare the performance of the MLE rule-of-thumb estimator and conventional kernel density estimators on artificial data with data properties that are varied in a controlled fashion and on a number of representative real-world pattern recognition tasks, to gain a better understanding of the behaviour of these estimators in high-dimensional spaces and to determine whether these estimators are suitable for initialising the bandwidths of iterative ML bandwidth estimators in high dimensions. We find that there are several regularities in the relative performance of conventional kernel density estimators across different tasks and dimensionalities and that the Silverman rule-of-thumb bandwidth estimator performs reliably across most tasks and dimensionalities of the pattern recognition datasets considered, even in high-dimensional feature spaces. Based on this empirical evidence and the intuitive theoretical motivation that the Silverman estimator optimises the asymptotic mean integrated squared error (assuming a Gaussian reference distribution), we select this estimator to initialise the bandwidths of the iterative ML kernel bandwidth estimators compared in our simulation studies. We then perform a comparative simulation study of the newly introduced iterative MLE estimators and other state-of-the-art iterative ML estimators on a number of artificial and real-world high-dimensional pattern recognition tasks. We illustrate with artificial data (guided by theoretical motivations) under what conditions certain estimators should be preferred and we empirically confirm on real-world data that no estimator performs optimally on all tasks and that the optimal estimator depends on the properties of the underlying density function being estimated. We also observe an interesting case of the bias-variance trade-off where ML estimators with fewer parameters than the MLE estimator perform exceptionally well on a wide variety of tasks; however, for the cases where these estimators do not perform well, the MLE estimator generally performs well. The newly introduced MLE kernel bandwidth estimators prove to be a useful contribution to the field of pattern recognition, since they perform optimally on a number of real-world pattern recognition tasks investigated and provide researchers and practitioners with two alternative estimators to employ for the task of kernel density estimation. / PhD (Information Technology), North-West University, Vaal Triangle Campus, 2014

With the advent of the internet and advances in computing power, the collection of very large high-dimensional datasets has become feasible { understanding and modelling high-dimensional data has thus become a crucial activity, especially in the field of pattern recognition. Since non-parametric density estimators are data-driven and do not require or impose a pre-defined probability density function on data, they are very powerful tools for probabilistic data modelling and analysis. Conventional non-parametric density estimation methods, however, originated from the field of statistics and were not originally intended to perform density estimation in high-dimensional features spaces { as is often encountered in real-world pattern recognition tasks. Therefore we address the fundamental problem of non-parametric density estimation in high-dimensional feature spaces in this study. Recent advances in maximum-likelihood (ML) kernel density estimation have shown that kernel density estimators hold much promise for estimating nonparametric probability density functions in high-dimensional feature spaces. We therefore derive two new iterative kernel bandwidth estimators from the maximum-likelihood (ML) leave one-out objective function and also introduce a new non-iterative kernel bandwidth estimator (based on the theoretical bounds of the ML bandwidths) for the purpose of bandwidth initialisation. We name the iterative kernel bandwidth estimators the minimum leave-one-out entropy (MLE) and global MLE estimators, and name the non-iterative kernel bandwidth estimator the MLE rule-of-thumb estimator. We compare the performance of the MLE rule-of-thumb estimator and conventional kernel density estimators on artificial data with data properties that are varied in a controlled fashion and on a number of representative real-world pattern recognition tasks, to gain a better understanding of the behaviour of these estimators in high-dimensional spaces and to determine whether these estimators are suitable for initialising the bandwidths of iterative ML bandwidth estimators in high dimensions. We find that there are several regularities in the relative performance of conventional kernel density estimators across different tasks and dimensionalities and that the Silverman rule-of-thumb bandwidth estimator performs reliably across most tasks and dimensionalities of the pattern recognition datasets considered, even in high-dimensional feature spaces. Based on this empirical evidence and the intuitive theoretical motivation that the Silverman estimator optimises the asymptotic mean integrated squared error (assuming a Gaussian reference distribution), we select this estimator to initialise the bandwidths of the iterative ML kernel bandwidth estimators compared in our simulation studies. We then perform a comparative simulation study of the newly introduced iterative MLE estimators and other state-of-the-art iterative ML estimators on a number of artificial and real-world high-dimensional pattern recognition tasks. We illustrate with artificial data (guided by theoretical motivations) under what conditions certain estimators should be preferred and we empirically confirm on real-world data that no estimator performs optimally on all tasks and that the optimal estimator depends on the properties of the underlying density function being estimated. We also observe an interesting case of the bias-variance trade-off where ML estimators with fewer parameters than the MLE estimator perform exceptionally well on a wide variety of tasks; however, for the cases where these estimators do not perform well, the MLE estimator generally performs well. The newly introduced MLE kernel bandwidth estimators prove to be a useful contribution to the field of pattern recognition, since they perform optimally on a number of real-world pattern recognition tasks investigated and provide researchers and practitioners with two alternative estimators to employ for the task of kernel density estimation. / PhD (Information Technology), North-West University, Vaal Triangle Campus, 2014

As the Internet continues to evolve, without providing any performance guarantees or explicit feedback to applications, the only way to infer the state of the network and to dynamically react to congestion is through end-to-end measurements. The emph{available bandwidth} (avail-bw) is an important metric that characterizes the dynamic state of a network path. Its measurement has been the focus of significant research during the last 15 years. However, its estimation remained elusive for several reasons. The main contribution of this thesis is the development of the first estimation methodology for the avail-bw in a network path using end-to-end measurements. In more detail, our first contribution is an end-to-end methodology, called SLoPS, to determine whether the avail-bw is larger than a given rate based on the sequence of one-way delays experienced by a periodic packet stream. The second contribution is the design of two algorithms, based on SLoPS, to estimate the mean and the variation range, respectively, of the avail-bw process. These algorithms have been implemented in two measurement tools, referred to as PathLoad and PathVar. We have validated the accuracy of the tools using analysis, simulation, and extensive experimentation. Pathload has been downloaded by more than 6000 users since 2003. We have also used PathVar to study the variability of the avail-bw process as a function of various important factors, including traffic load and degree of multiplexing. Finally, we present an application of avail-bw estimation in video streaming. Specifically, we show that avail-bw measurements can be used in the dynamic selection of the best possible overlay path. The proposed scheme results in better perceived video quality than path selection algorithms that rely on jitter or loss-rate measurements.