Aggregated Multicast for Scalable QoS Multicast Provisioning

07/2001; DOI: 10.1007/3-540-45400-4_20
Source: CiteSeer

ABSTRACT IP multicast suffers from scalability problem with the number of concurrently active multicast groups, while scalability of QoS multicast is even further from being solved. In this paper, we propose an approach to reduce multicast forwarding state and provision multicast with QoS guarantees. In our approach, multiple groups are forced to share a single delivery tree. We discuss the advantages and some implementation issues of our approach, and conclude that it is feasible and promising. We then describe how to use our approach to provision scalable QoS multicast. Finally, we define metrics to quantify state reduction and use simulations to show how our scheme achieves state reduction. These initial simulation results suggest that our method can reduce multicast state significantly.

  • [Show abstract] [Hide abstract]
    ABSTRACT: The main question of aggregated Multicast is how to match group to tree perfectly. In this paper, we present a new group-tree matching algorithm which is based on group-member matched group-tree matching algorithm, and here it is called GTBA algorithm. It is presented on the base of analyzing and concluding the traditional group-tree matching algorithms' advantages and disadvantages. The GTBA algorithm can realize the multicast trees merging in the presence of saving the network bandwidth resources. Due to the dynamic of Multicast, the GTBA algorithm can achieve refresh the forwarding table quickly and consequently improves the forwarding efficiency in the network. Finally, we prove that the validity of the GTBA algorithm by simulation.
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: IP Multicasting suffers from state scalability as the number of concurrently active multicast groups increases, since it requires a router to keep forwarding state for every multicast tree passing through it. In multicast provisioning in a Diff_Serv domain the condition is even worse, since, in addition to multicast state, the requested PHB for each group need to be kept at the routers. In this paper we present an architecture called Differentiated Services-aware-Aggregated Multicast (DSAM), and we examine how we can use aggregated multicast to support multicasting in a Diff_Serv supported MPLS network. Proposed group-tree mapping algorithm, takes into account the PHB requirement for each group and can find a proper aggregated tree. DSAM is scalable and reduces multicast state and processing and control overheads in the core routers.
  • [Show abstract] [Hide abstract]
    ABSTRACT: Teaching Software Engineering (SE) based on “real-world” projects engages students with practical application of software engineering concepts-students develop a deeper interest in the project deliverables while they acquire the skills of critically analyzing the problem and determining the best course of action. It can be challenging to find and maintain a reliable stream of suitable software projects that match learning outcomes, technical scope, and academic calendar of an SE class. On the other hand, a typical university campus has many nonComputer Science (CS) technology classes that require their students to study, understand, and evaluate existing software applications in specific areas. With purposeful coordination, these non-CS technology classes can serve as effective source of real projects for SE classes. This paper describes our experience of collaborating with the Education Program in their Technology in Education course. While we encountered some challenges, our experience has demonstrated that it is indeed mutually beneficial and rewarding for students in both courses. We offer recommendations on choosing non-CS technology classes and logistical guidelines to ensure the success of such collaborations.
    Frontiers in Education Conference (FIE), 2012; 01/2012


Available from