• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2004.10a
• /
• pp.640-643
• /
• 2004

As the size of a network increases, the network complexity and maintenance cost is high. In literatures, many cost factors of multistage interconnection networks were intensively investigated. However, in some parts, these types of cost definition are not satisfactory enough since they don't relate the cost measure to technology closely and also fail to incorporate exact tradeoffs in designing the network. In this paper, we propose some cost effective measures which combine the cost factors with performance-related reliability measures, and compare several networks for a guidance to select a certain case depending on the principal concerns for use.

• Proceedings of the IEEK Conference
• /
• summer
• /
• pp.325-329
• /
• 2004

Linear size expanders have been studied in many fields for the practical use, which is possibility to connect large numbers of device chips in parallel communication systems. One major limitation on the efficiency of parallel computer designs has been the prohibitively high cost of parallel communication between processors and memories. Linear size expanders can be used to construct theoretically optimal interconnection networks. In current, the defined constructions have large constant factors, thus rendering them impractical for reasonable sized networks. This paper presents an improvement on constructing concentrators using an $(n,\;k,\;2rs/(r^2-s^2))expander,$ which realizes the reduction of the size in a superconcentrator by a constant factor.

• ETRI Journal
• /
• v.45 no.1
• /
• pp.60-74
• /
• 2023

Integration of the machine learning (ML) technique in all-optical networks can enhance the effectiveness of resource utilization, quality of service assurances, and scalability in optical networks. All-optical multistage interconnection networks (MINs) are implicitly designed to withstand the increasing highvolume traffic demands at data centers. However, the contention resolution mechanism in MINs becomes a bottleneck in handling such data traffic. In this paper, a select list of ML algorithms replaces the traditional electronic signal processing methods used to resolve contention in MIN. The suitability of these algorithms in improving the performance of the entire network is assessed in terms of injection rate, average latency, and latency distribution. Our findings showed that the ML module is recommended for improving the performance of the network. The improved performance and traffic grooming capabilities of the module are also validated by using a hardware testbed.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.48 no.2
• /
• pp.6-13
• /
• 2011

This paper deals with the interconnection problem in ad-hoc networks or sensor networks, where relay nodes are deployed additionally to form connections between given nodes. This problem can be reduced to a NP-hard problem. The nodes of the networks, by applications or geographic factors, can be deployed densely in some areas while sparsely in others. For such a case one can make an approximation scheme, which gives shorter execution time, for the additional node deployments by ignoring the interconnections inside the dense area of nodes. However, the case is still a NP-hard, so it is proper to establish a polynomial time approximation scheme (PTAS) by implementing a dynamic programming. The analysis can be made possible by an elaboration on making the definition of the objective function. The objective function should be defined to be able to deal with the requirement incurred by the substitution of the dense area with its abstraction.

• The KIPS Transactions:PartA
• /
• v.15A no.4
• /
• pp.189-198
• /
• 2008

In this paper, we prove mesh-like networks can be embedded into Petersen-Torus(PT) networks. Once interconnection network G is embedded in H, the parallel algorithm designed in Gcan be applied to interconnection network H. The torus is embedded into PT with dilation 5, link congestion 5 and expansion 1 using one-to-one embedding. The honeycomb mesh is embedded into PT with dilation 5, link congestion 2 and expansion 5/3 using one-to-one embedding. Additional, We derive average dilation. The embedding algorithm could be available in both wormhole routing system and store-and-forward routing system by embedding the generally known Torus and honeycomb mesh networks into PT at 5 or less of dilation and congestion, and the processor throughput could be minimized at simulation through one-to-one.

• Journal of Internet Computing and Services
• /
• v.21 no.5
• /
• pp.9-20
• /
• 2020

This study redefines the networks' connection behaviors and the terms confusion over the settlement in Netflix-SK Broadband's dispute through domestic and foreign legal references. Conflict parties, academics and the media use the terms "interconnection fee" or "Access fee" without uniformity, and in some cases mixes for strategic purposes. The use of different terms for the same phenomenon (or vice versa) has a high need for research in that it makes it difficult to reach a unified approach to the problem, to discuss it productively and rationally, and, moreover, to resolve disputes. Therefore, this study cross-referenced/analyzed terms related to network utilization and connectivity, namely "Use", "Access", "Interconnection" and thus cost-related terms as a counter-pay. In addition, it suggests that interconnection fees and access fees should be used separately, and allows them to function as a starting point in resolving future ICT sector issues. As a result of this study, the price against the network access/use between Netflix and SK Broadband is access fee or retail price, and proposes to be used uniformly in the term "interconnection fee" only for fees incurred in interconnection between ISPs that possess or operate networks.

• Journal of KIISE:Computer Systems and Theory
• /
• v.32 no.8
• /
• pp.393-398
• /
• 2005

In interconnection networks, a Hamiltonian path has been utilized in many applications such as the implementation of linear array and multicasting. In this paper, we consider the Hamiltonian properties of mesh networks which are used as the topology of parallel machines. If a network is strongly Hamiltonian laceable, the network has the longest path joining arbitrary two nodes. We show that a two-dimensional mesh M(m, n) is strongly Hamiltonian laceabie, if $m{\geq}4,\;n{\geq}4(m{\geq}3,\;n{\geq}3\;respectively)$, and the number of nodes is even(odd respectively). A mesh is a spanning subgraph of many interconnection networks such as tori, hypercubes, k-ary n-cubes, and recursive circulants. Thus, our result can be applied to discover the fault-hamiltonicity of such networks.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.30 no.4
• /
• pp.151-161
• /
• 2005

For Internet service providers (ISPs), there are three common types of interconnection agreements : private peering, public peering and transit. One of the most important problems for a single ISP is to determine which other ISPs to interconnect with, and under which agreements. The problem can be then to find a set of private peering providers, transit providers and Internet exchanges (IXs) when the following input data are assumed to be given : a set of BGP addresses with traffic demands, and a set of potential service providers (Private peering/transit providers and IXs) with routing information, cost functions and capacities. The objective is to minimize the total interconnection cost. We show that the problem is NP-hard, give a mixed-integer programming model, and propose a heuristic algorithm. Computational experience with a set of test instances shows the remarkable performance of the proposed algorithm of rapidly generating near-optimal solutions.

• Journal of information and communication convergence engineering
• /
• v.20 no.3
• /
• pp.160-165
• /
• 2022

This paper proposes an efficient method of reconstructing interconnections when the terminals of each plane change in real-time situations where randomly divided planes are interconnected. To connect all terminals when the terminals of each plane are changed, we usually reconstruct the interconnections between all terminals. This ensures a minimum connection length, but it takes considerable time to reconstruct the interconnection for the entire terminal. This paper proposes a solution to obtain an optimal tree close to the minimum spanning tree (MST) in a short time. The construction of interconnections has been used in various design-related areas, from networks to architecture. One of these areas is an ad hoc network that only consists of mobile hosts and communicates with each other without a fixed wired network. Each host of an ad hoc network may appear or disappear frequently. Therefore, the heuristic proposed in this paper may expect various cost savings through faster interconnection reconstruction using the given information in situations where the connection target is changing.

• Journal of KIISE:Computer Systems and Theory
• /
• v.32 no.5
• /
• pp.196-208
• /
• 2005

In multiprocessor systems, interconnection network design is critical for overall system performance. Popular interconnection networks, which are generally considered, are meshes, rings, and hierarchical rings. In this paper, we propose (')Torus Ring('), which is a modified version of hierarchical ring. Torus Ring has the same complexity as the hierarchical rings, but the only difference is the way it connects the local rings. It has an advantage over the hierarchical rings when the destination of a packet is the neighbor local ring in the reverse direction. Though the average number of hops in Torus Ring is equal to that of the hierarchical rings when assuming the uniform distribution of each transaction, the benefits of the number of hops are expected to be larger because of the spatial locality in the real environment of parallel programming. In the simulation results, latencies in the interconnection network are reduced by up to 19$\%$, and the execution times are reduced by up to 10$\%$.