Thesis (M.Ing. (Computer and Electronical Engineering))--North-West University, Potchefstroom Campus, 2009.

Network topologies play a vital role in chip design; they largely determine the cost of the network and significantly impact performance in many-core architectures. We propose a novel set of on-chip networks, dodecs, and illustrate how they reduce network diameter with randomized low-radix router connections. In addition, we design an adaptive routing algorithm for dodec networks to achieve high throughput. By introducing randomness, dodec networks exhibit more uniform message latency. By using low-radix routers, dodec networks simplify the router microarchitecture and attain 20% area and 22% power reduction compared to mesh routers while delivering the same overall application performance for PARSEC. We compare our dodec network to alternative low-radix network topologies and show that at the same cost, dodec networks increase the throughput up to 50% while reducing average latency by 10% compared to a mesh.

Network topologies play a vital role in chip design; they largely determine the cost of the network and significantly impact performance in many-core architectures. We propose a novel set of on-chip networks, dodecs, and illustrate how they reduce network diameter with randomized low-radix router connections. In addition, we design an adaptive routing algorithm for dodec networks to achieve high throughput. By introducing randomness, dodec networks exhibit more uniform message latency. By using low-radix routers, dodec networks simplify the router microarchitecture and attain 20% area and 22% power reduction compared to mesh routers while delivering the same overall application performance for PARSEC. We compare our dodec network to alternative low-radix network topologies and show that at the same cost, dodec networks increase the throughput up to 50% while reducing average latency by 10% compared to a mesh.

Recent advances in methods to increase network utilization have lead to the introduction of a relatively new method called Network Coding. Network Coding is a method that can reduce local congestion in a network by combining information sent over the network. It is commonly researched in the information theory field after it was first introduced by Ahlswede et al in 2000. Network Coding was proven in 2003, by Koetter & Medard to be the only way to achieve the throughput capacity defined by the Min cut Max flow theorem of Shannon. It was applied deterministically in wired networks and randomly in wireless networks. Random Network Coding however requires a lot of overhead and may cause possible delays in the network. We found that there is an open question as to determine where in a wireless network, Network Coding can be implemented. In this thesis we propose to find opportunities for the implementation of Network Coding, by searching for known deterministic Network Coding topologies in larger Networks. Because a known topology is used, we will then also know how Network Coding should be implemented. This method of finding opportunities for the implementation of Network Coding using topology can be combined with a routing algorithm to improve the utilization of a wireless network. We implemented our method on three different topologies and searched 1000 random networks for the presence of these topologies. We found that these topologies occurred frequently enough to make our method a viable method of finding opportunities for the implementation of Network Coding. / Thesis (M.Ing. (Computer and Electronical Engineering))--North-West University, Potchefstroom Campus, 2009.

Recent advances in methods to increase network utilization have lead to the introduction of a relatively new method called Network Coding. Network Coding is a method that can reduce local congestion in a network by combining information sent over the network. It is commonly researched in the information theory field after it was first introduced by Ahlswede et al in 2000. Network Coding was proven in 2003, by Koetter & Medard to be the only way to achieve the throughput capacity defined by the Min cut Max flow theorem of Shannon. It was applied deterministically in wired networks and randomly in wireless networks. Random Network Coding however requires a lot of overhead and may cause possible delays in the network. We found that there is an open question as to determine where in a wireless network, Network Coding can be implemented. In this thesis we propose to find opportunities for the implementation of Network Coding, by searching for known deterministic Network Coding topologies in larger Networks. Because a known topology is used, we will then also know how Network Coding should be implemented. This method of finding opportunities for the implementation of Network Coding using topology can be combined with a routing algorithm to improve the utilization of a wireless network. We implemented our method on three different topologies and searched 1000 random networks for the presence of these topologies. We found that these topologies occurred frequently enough to make our method a viable method of finding opportunities for the implementation of Network Coding. / Thesis (M.Ing. (Computer and Electronical Engineering))--North-West University, Potchefstroom Campus, 2009.

Thesis (Ph. D.)--Rutgers University, 2007. / "Graduate Program in Electrical and Computer Engineering." Includes bibliographical references (p. 140-144).

Thesis (M.S.)--California Polytechnic State University, 2009. / Title from PDF title page; viewed on August 11, 2009. Major professor: Taufik, Ph.D. "Presented to the faculty of the College of Engineering, California Polytechnic State University, San Luis Obispo." "In partial fulfillment of the requirements for the degree [of] Master of Science in Electrical Engineering." "June 11, 2009." Includes bibliographical references (p. 58-59). Will also be available on microfiche.

Thesis (Ph. D.)--Washington State University, August 2007. / Includes bibliographical references (p. 125-133).

Bibliography: p. 17. / "May 1984." / NSF-ECS-8310698

The objective of this thesis is creating a design of a redundant secure VPN WAN network for Medical Emergency Service of Pardubice region. The starting point of this thesis is the analysis of the current state of the corporate computer network, which was evaluated as not satisfying. The result is a design of WAN network, which satisfies the requirements of the investor. The solution also includes the project of implementation and financial calculation of the project.