Ki-Sung Yu

Network Stability Assessment Using the Number of Tree Adjacent to an Articulation Node

In this study, we examine the stability assessment of the current network by using the number of trees generated when an articulation node within the network becomes functional. As a consequence, we will be able to design networks with more stability using the same amount of resources. Previous studies assessed the stability of a network by using the number of nodes adjacent to a broken node but they were inapplicable to complex networks with more than two broken nodes. The network stability assessment method proposed in this study will be applicable to complex networks with two or more broken nodes and will offer a more detailed assessment by considering the distribution of network risk factors.

Optimum Numbers of Tunable Wavelength Converters and Internal Wavelengths in the Optical Packet Switch with Shared FDL Buffer

In an effort to reduce switch cost, we, for the first time, present the optimum numbers of tunable wavelength converters (TWCs) and internal wavelengths required for contention resolution of asynchronous and variable length packets, in the optical packet switch (OPS) with the shared fiber delay line (FDL) buffer. To optimize TWCs and internal wavelengths related to an OPS design cost, we proposed a scheduling algorithm for the limited TWCs and internal wavelengths. For three TWC alternatives (not shared, partially shared, and fully shared cases), the optimum numbers of TWCs and internal wavelengths to guarantee minimum packet loss are evaluated to prevent resource waste. Under a given load, TWCs and internal wavelengths could be significantly reduced, guaranteeing the same packet loss as the performance of an OPS with full TWCs and internal wavelengths

Design of network aware resource allocation system for grid applications

Grid Network applications that need to achieve reliable end-to-end performance typically make use of either reservations or adaptation. However, both techniques also have serious limitations, particularly when dealing with high-bandwidth, dynamic flow as like Grid. Fixed-capability reservations tend to be wasteful of resources and it cannot permit dynamic adaptation in application performance when resource management policies are changed. Adaptation scheme inevitably fails when congestion reduces available resources below acceptable limits. In this paper, we describe an approach to QoS that combines features of both reservation and adaptation to address the difficulties above noted. In addition, we propose extended GARA and advanced reservation and reserved resource saving scheme that provide more effective QoS.

Recovery Schemes for Fast Fault Recovery on GMPLS Network

In this research, we proposed two types of fault recovery schemes designed by combining the protection and the restoration methods in order to increase the survivability of GMPLS networks. The first method is a transformation of the 1+1 protection method where it consumes more resources than the 1+1 protection method but has a faster fault recovery speed. The second method is a combination of the 1+1 protection method and the restoration method and can enhance the performance after recoveries. Although the two methods proposed in this paper consume more resources than the ones before, they have outstanding recovery time and performance after recoveries.

Improved fast handover protocol using HMIPv6 based on IEEE 802.16e network

Since the Wibro service realize mobile network on the public, it has been considered that it is not enough to support real time service at vehicular speed. The standard Fast Mobile IPv6 Handover Protocol using HMIPv6 may guarantee seamless service as long as the Mobile Node moves in the same domain MAP however it does not regard fast handover over inter-MAP domain. Thus Macro Mobility Handover in HMIPv6 was proposed to reduce handover latency in inter-MAP domain. But it is still not enough to support real-time service. So we propose an Improved FH- MIPv6 over 802.16e network to reduce the overall handover latency. We embedded Layer 3 handover messages of the FHMIPv6 into the Layer 2 handover messages. So the MN is able to finish overall handover procedure earlier in our scheme. The numerical results show the performance of IFHMIPv6 has improved 32% in comparison with FHMIPv6.

Resilient data delivery in delay bounded overlay multicast

Overlay multicast has been proposed as an alternative scheme to provide one-to-many or many-to-many data delivery on Internet. However, since data delivery is entirely dependent on replications on each group member, if one member node fails, none of its children receives data. Furthermore, the higher the failed nodes level is, the more the dead nodes occur. In this paper, we give a detailed framework for resilience in overlay multicast. Unlike previous efforts in providing reliable data delivery, our scheme constructs reliable overlay data delivery tree (DDT) and adaptively reconstructs it based on the number of missed packets while guaranteeing end-to-end delay bound. Through practical simulation results, we analyzed packet delivery ratio, control overhead, end-to-end delay and number of replicated packets