Byung Ho Yae

Reduction of Handover Latency Using MIH Services in MIPv6

In this paper, we propose an enhanced handover mechanism with new additional primitives and parameters to the media independent handover (MIH) services defined in the IEEE 802.21. The proposed scheme can reduce handover latency for mobile IPv6 (MIPv6) by removing the router discovery time. Moreover, when the proposed mechanism is applied to the FMIPv6, we can increase the probability that the FMIPv6 can be performed in predictive mode by reducing the handover initiation time, thereby we can reduce the expected handover latency in the FMIPv6. In addition, with the proposed scheme, we can design the network cost-effectively by reducing coverage overlap between adjacent cells because the handover initiation time in the FMIPv6 is decreased

Enhanced fast handover mechanism using MIH services in MIPv6

In this paper, we propose an enhanced handover mechanism with new additional primitives and parameters to the media independent handover (MIH) services as defined in IEEE 802.21. The proposed scheme can reduce handover latency for MIPv6 (mobile IPv6) by removing the router discovery time. Moreover, when the proposed mechanism is applied to a FMIPv6 (fast handovers for MIPv6), we can increase the probability that FMIPv6 can be per-formed in a predictive mode by reducing the handover initiation time. As a result, we can get the reduced handover latency in FMIPv6. In addition, with the proposed scheme, we can design the cost-effective network by reducing cover-age overlap between adjacent cells. This is due to the decreased handover initiation time in FMIPv6. From simulation results, it is proven that the proposed scheme can provide a higher goodput for UDP and TCP because it has less service disruption during handovers.

A practical RWA algorithm based on lookup table for edge disjoint paths

Routing and wavelength assignment (RWA) problem is an important issue in optical transport networks based on wavelength division multiplexing (WDM) technique. It is typically solved using a combination of linear programming and graph coloring, or path selection based graph algorithms. Such methods are either complex or make extensive use of heuristics. In this paper we propose a novel and efficient approach which basically obtains the maximum edge disjoint paths (EDPs) for each source-destination demand pair. And those EDPs obtained are stored in Lookup Table and used for the update of weight matrix. Routes are determined in order by the weight matrix for the demand set. The comprehensive computer simulation shows that the proposed algorithm uses similar or fewer number of wavelengths with significantly less execution time than bounded greedy approach (BGA) for EDP [2] which is currently known to be effective in practice.

Implementation Issues for IPv6 Software Forwarding in Routing System

This paper considers software forwarding intending as the short solution for deploying IPv6 feature in the routing system which has already IPv4 hardware forwarding function. Linux kernel forwards IPv6 packet through looking up the routing table being made from upper software blocks. Essentially we had to change each interface setup, forwarding base management and neighbor information control scheme. Considering the amount of current IPv6 traffic, the performance of IPv6 software forwarding is acceptable and this work is prototype for deploying IPv6 hardware forwarding

Analysis framework of layer 1 VPNs using PCEMP

When path computation is done in the management systems for Layer 1 Virtual Private Networks (L1 VPNs), it would be important that resource information is synchronized between the management systems and the provider edge/provider (PE/P). This paper looks at such a scenario from the viewpoint of the path computation metric element protocol (PCEMP) that acts a generic computation model for path based metrics in large multidomain or multilayer networks. In this paper, we show a preliminary analysis framework of L1 VPN architectures based on path computation element (PCE) architecture [A. Farrel et al. (2004)] and the PCEMP protocol [J. K. Choi et al. (2004)]. The analysis is discussed in the context of trajectory splitting simulation techniques based on the concept of direct probability redistribution (DPR), the fundamentals of which are discussed in [Y. Akin et al. (2001), M. Devetsikiotis et al. (1993)]. As the PCE based network environments are protocol driven, we show how the design and architecture features of the protocol and the consequential network may help in resolving scalability issues for large scale inter-domain routing and service provisioning in L1 VPNs. The novelty of the approach lies in the fact that this simulation technique has not been used in the context of protocol driven architectures before

QoS class mapping mechanism at the access gateway

This paper is concerned about an optimum access control algorithm to guarantee quality of service over an access networks. First, protocol and network architecture after introducing a motive and circumstance is mentioned. And there is some analysis of the queuing model for the access network. Finally the author concluded the effect and performance of this mechanism