Physical Sciences and Mathematics Commons^™

Modeling The Multicast Address Allocation Problem, Daniel Zappala, Chris Gauthierdickey, Virginia Lo

Faculty Publications

To support IP multicast, domains must assign a unique multicast address to each application from a limited, globally-shared address space. We examine the performance of several classes of address allocation algorithms withln the context of the MASC architecture. This study is the first of its kind to model the generalized multicast address allocation problem and consider non-contiguous allocation algorithms. We find that prefix-based allocation outperforms our non-contiguous algorithm, despite the apparent advantages of non-contiguous allocation. We also verify the benefits of using worst-fit for new allocations.

A Theoretical Framework For The Multicast Address Allocation Problem, Daniel Zappala, Chris Gauthierdickey, Virginia Lo, Timothy Singer

Faculty Publications

The multicast address allocation problem requires Internet domains to allocate unique addresses to multicast applications from a globally-shared space. We develop a theoretical framework for multicast allocation algorithms that is influenced by subcube allocation in hypercube computer systems. Based on this framework we derive complexity results for the address allocation problem and describe several new allocation algorithms that use a hypercube model for address representation.

Performance Evaluation Of Path Searching Heuristics For Multicast Qos Routing, Daniel Zappala, Dayi Zhou

Faculty Publications

Quality of Service routing for multicast enables a group member to find and install a branch of the multicast tree that can meet its QoS constraints. The most promising protocols in this area use receiver-oriented path searching heuristics to find feasible routes. In this paper, we examine the performance of the path searching heuristics used by these protocols, to determine which techniques are the most effective. We find that several low-overhead path searching heuristics are effective, and that generic path searching can work as well as targeted, QoS-specific searching.

Scalable And Reliable File Transfer For Clusters Using Multicast., Hardik Dikpal Shukla

Electronic Theses and Dissertations

A cluster is a group of computing resources that are connected by a single computer network and are managed as a single system. Clusters potentially have three key advantages over workstations operated in isolation—fault tolerance, load balancing and support for distributed computing.

Information sharing among the cluster’s resources affects all phases of cluster administration. The thesis describes a new tool for distributing files within clusters. This tool, the Scalable and Reliable File Transfer Tool (SRFTT), uses Forward Error Correction (FEC) and multiple multicast channels to achieve an efficient reliable file transfer, relative to heterogeneous clusters. SRFTT achieves scalability …