| The aim of this dissertation is to develop effective dynamic bandwidth management algorithms to meet the multiple quality of service (QoS) requirements of the various applications utilizing a telecommunication network. In recent years, growing processing power and bandwidth have stimulated a new breed of applications, such as IP-based Private Branch Exchange, video on demand and video conferencing. These new applications require more innovative traffic management algorithms to meet multiple QoS requirements, and future applications promise to be even more challenging. One of the main obstacle to improving QoS is congestion at network nodes, which results in cell loss, delay and traffic fluctuation. It is believe that predicting the bandwidth at the output of each network node is the key to reducing such congestion and, thus, improving QoS. The proposed method is compatible with current technologies such as Voice Over IP (VoIP), Multi-protocol Label Switching (MPLS) and asynchronous transfer mode (ATM). The novelty of this approach hinges on its exploitation of the discrete wavelet transform (DWT) of the network traffic. Recent studies have shown that while network traffic consists of both long- and short-term dependence, its wavelet coefficients are short-term dependent only. Hence, significant advantages may be gained by employing legacy traffic models in the wavelet domain. The DWT decomposes the traffic into a low-frequency component and its high frequency components—the details. The low-frequency component of the traffic, are significant for long-term traffic behavior and, hence, for bandwidth allocation. The high-frequency components are significant in the short-term behavior of the traffic and, hence, for buffer-allocation. |
