• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37B no.10
• /
• pp.877-888
• /
• 2012

Network virtualization technology is a crucial research issue of Future Internet which pursues a service-oriented architecture so-called NaaS (Network as a Service) or SDN (Software Defined Network). Network virtualization is expected to play an important role in Future Internet researches as a network testbed technology which enables innovative protocols to be experimented independently on a common testbed environment. We propose a control framework in order to provide user defined topology and bandwidth services with network virtualization and to separate and manage multiple-user traffics in a small and medium scale - single domain research network. The proposed framework (VIMS; Virtual network Integrated control and Management System) supports testbed expansions without any changes of heterogeneous virtual network support equipments through accommodation of each equipment's control plane. The framework shows a feasibility through applied to KOREN and we describe the differences and further study directions for improvement the framework comparing with GENI control framework.

• The Journal of the Korea Contents Association
• /
• v.10 no.4
• /
• pp.55-65
• /
• 2010

We present a method for increasing network lifetime without link failure due to lack of battery capacity of nodes in wireless ad-hoc networks with low mobility. In general, a node with larger remaining battery capacity represents the one with lesser traffic load. Thus, a modified AODV routing protocol is proposed to determine a possible route by considering a remaining battery capacity of a node. Besides, the total energy consumption of all nodes increase rapidly due to the huge amount of control packets which should be flooded into the network. To reduce such control packets efficiently, a source node can store information about alternative routes to the destination node into its routing table. When a link failure happens, the source node should retrieve the route first with the largest amount of the total remaining battery capacity from its table entries before initiating the route rediscovery process. To do so, the possibility of generating unnecessary AODV control packets should be reduced. The method proposed in this paper increases the network lifetime by 40% at most compared with the legacy AODV and MMBCR.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.30 no.8B
• /
• pp.514-524
• /
• 2005

In optical burst switching(OBS) networks, the ingress edge router assembles packets in the same class queue into the appropriate size of burst. A burst control packet(BCP) is generated for channel reservation of corresponding data burst and sent earlier than the corresponding data burst with an offset time. Offset time is determined considering the number of hops from source to destination and the required quality of service(QoS). After offset time, the burst data is passed through tile pre-configured optical switches without any O/E/O conversion. But a failure in OBS networks may lead to the loss of bursts until the ingress nodes receive the failure indication signal. This results in a significant degradation in QoS. Therefore, in this paper, we propose a fault-tolerant node architecture based on sub-carrier multiplexing to reduce the effects of failure in OBS networks. The Performance of the proposed fault-tolerant node architecture exhibits considerable improvement as compared with the previous ones.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.21 no.3
• /
• pp.664-674
• /
• 1996

In high speed multi-wavelength networks, retransmission overhead due to desination conflict or control packet collision is one factor of performance degration because signal prpagation delay is much larger than the transmission time of data packet. In this paper, an efficient WDMA protocol with a collision avoidance mechanism is proposed for high speed WDM single-hop network with a passive star topology. In proposed protocol, each node has cource queues and routing table to store souting informatio. This architecture makes is possible to avoid any kind of collision when a node reserves the channel to transmit a data packet. High system thoughput and channel utilization can be achieved by proposed protocol since there are no discarded packets caused by any collision at transmission time. The performance of proposed protocol is evaluated in term of throughput and delay with variations in offered load. Simulation results show that the proposed protocol has superior performance to convertional protocols under nonuniform traffic as well as uniform traffic.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.36 no.12B
• /
• pp.1601-1610
• /
• 2011

This paper discusses standardization activities, requirements and enabling technologies for 100G Ethernet and 100G OTN. The need of 100Gbps transport capacity has been gaining greater interest from service providers and carrier vendors. Moreover, optical transport networks based on OTN/DWDM are changing their properties to apply Ethernet traffic which is dramatically increasing. We realize and experimentally demonstrate OTU4 framer with commercial FPGA. The key features of the realized OTU4 framer are parallel signal processing function, multi-lane distribution function, GMP function and FEC function. The realized OTU4 framer has the large signal processing capacity of 120Gbps, which allows to transport about 120Gbps client signals such as $12{\times}10G$ Ethernet and $3{\times}40G$ Ethernet. The realized OTU4 framer has the advantages to quickly adjust to changing markets and new technologies by using commercial FPGA instead of ASIC.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.1B
• /
• pp.16-23
• /
• 2009

This paper investigates how many optimal routes can be established in terms of a given wireless mesh routing metric. Although many of routing metrics have been devised to precisely derive the wireless link quality in mesh, most (if not all) metrics have not been evaluated their optimality along with routing protocols. We consider stateof-the-art routing metrics and a widely accepted routing protocol in order to observe the optimality of established routes varying the number of source nodes. Also, we propose a unidirectional routing to deal with possible link asymmetry feature in wireless links. Through comparative simulation evaluations, we show that the portion of optimally established routes becomes less as the network traffic load increases, regardless of employed metrics, network topologies, and routing protocols.

• Journal of Digital Contents Society
• /
• v.14 no.2
• /
• pp.191-197
• /
• 2013

To complete the "anywhere" mobile service there should not be shaded areas. However, it is never economical to deploy BS's (Base Stations) in a low population density area due to the low resource utilization. We propose a design technique to install RS's(Relay Stations) which are cheaper than BS's and find the condition in which the proposed technique is acceptible. The proposed design technique aims to allocate the frequency and time resources to RS's to minimize the transmission rate degradation due to interferences between RS's as well as to maximize the parallelism in transmission. We showed by simulation that the proposed technique achieves cost benefits when the expected traffic is less than 20.32 percent of the total BS capacity. The proposed technique is compatible with the IEEE 802.16j, thus can be extended to multi-hop configuration.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.4 no.1
• /
• pp.17-25
• /
• 2000

ATM PON system is regarded as a solution of the next generation access network because it can cover various types of subscribers in a single platform. In this paper, we describe the merit of the ATM PON and the key protocols for its operation. Above all, the ranging is the most important protocol providing a mean of the time-division multiple access in the system. A problem in the protocol is that the window size, the time interval that ranging cell arrives, is too long. During the interval, user traffics cannot be send to the upstream, which makes the quality of service degraded. In this paper, we suggest a method to minimize the window size when the length of the optic fiber is known with some deviation. The window size can be reduced as 7 % of the conventional method when the deviation of the length is 1 Km.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.11
• /
• pp.2651-2657
• /
• 2014

It can be organized with PVRs and FC-AL for a high-quality TV content-sharing architecture. It is, however, not easy to configure a large community network with the FC-AL since it has a 7-bit address space with supporting up to 127 users. We, therefore, propose a novel scalable FC-AL multiple-loop architecture using shared disks, which does not need to use expensive and load-concentrated switches. In this article, our suggested architecture shows a superb startup delay, such as less than 20msec. In addition, it demonstrates outstanding scalability, such that the number of accomodable users increases almost linearly according to adding loops. Lastly, it reveals exceptional time-shifting hours, that is, which supports more than 140 hours with 1000 users.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.28 no.12B
• /
• pp.1066-1080
• /
• 2003

In this paper a scalable QoS routing scheme for distributed multimedia applications in a hierarchical wide area network is proposed. The problem of QoS routing is formulated as a multicriteria shortest path problem, known as NP-complete. The proposed hierarchical routing scheme consists of two phases. In Phase 1, every border node periodically pre-computes the QoS distance for the paths between every pair of border nodes in any level of domain hierarchy. This phase is independet of the QoS request from an application. In Phase II, distributed graph construction algorithm is performed to model the network as a graph by retrieving pre-computed QoS distances. The graph is constructed by the on-demand algorithm and contains a part of the network topology which is completely neglected or partially considered by existing routing schemes, thus maintaining more accurate topology information. By using retrieval approach rather than advertising one, no global QoS state information exchange among nodes is needed. In this Phase, distributed partition algorithm for QoS routing problem is also performed, thus eliminating virtual links on the hierarchically complete path.