Abstract

Multicasting has become increasingly important with the emergence of Internet-based applications such Internet protocol (IP) telephony, audio/video conferencing, distributed databases and software upgrading. IP Multicasting is an efficient way to distribute information from a single source to multiple destinations at different locations. One of the challenges the Internet is facing today is to keep the packet forwarding performance up with the skyrocketing demand for bandwidth. On the other hand, the MultiProtocol Label Switching (MPLS), which is an Internet Engineering Task Force (IETF) framework, combines the flexibility of layer 3 routing and layer 2 switching, which enhances network performance in terms of scalability, computational complexity, latency and control message overhead. Besides, MPLS offers a vehicle for enhanced network services such as Quality of Services (QoS)/Class of Service (CoS), Traffic Engineering and Virtual Private Networks (VPNs). In this thesis, we present a new Fair Share Policy (FSP), which is a traffic policing mechanism that utilizes Differentiated Services (DiffServ) to solve the problems of QoS and congestion control. We compare the QoS performance of IP and MPLS multicasting, given their particular constraints. In order to achieve the required QoS, different techniques of reliable multicasting are adapted, such as Forward Error Correction (FEC), Automatic Repeat Request (ARQ) or Hybrid FEC/ARQ with multicast or unicast repairs mechanisms so as to mitigate the effect of errors as well as packet loss. This reliable multicast is for both IP and MPLS platforms with Diffserv. Analytical and simulation models are suggested and employed. The results provide insights into the comparisons between IP multicast in MPLS networks using FSP and plain IP multicasting using the same policy when DiffServ is adopted and when reliable multicast is considered. This comparison will be based on the following QoS measures: total packet delay, delay jitter and residual packet loss probability. Analysis and simulation tools are used to evaluate our fair share policy (FSP) for different homogeneous (when all routers are identical in their capabilities) and heterogeneous (when routers have different capabilities) network scenarios.