IP망에서 트래픽 분할에 의한 다중경로 멀티캐스트 경로설정

박구현(Koohyun Park),신용식(Yong-Sik Shin) 한국정보과학회 2002 정보과학회논문지 : 정보통신 Vol.29 No.1

본 연구는 IP(Internet Protocol) 망에서 복수 트리경로에 의한 멀티캐스팅 방법을 제시한다. 트래픽 양에 대해 아크 비용이 비선형 함수이고 멀티캐스트 트래픽 수요가 대용량일 때, 단일의 트리경로 대신에 트래픽을 분할하여 복수개의 트리경로로 전송하는 것이 멀티캐스트 서비스 품질을 향상시킬 수 있다. 제시하는 멀티캐스팅 방법은 원하는 수준까지 트리경로를 추가하며, 주어진 트리경로 집합에 대해서 트래픽을 분할한다. 트래픽 분할을 위한 수학적 모형과 최적조건을 도출하였으며, 여러 개의 멀티캐스팅 트래픽을 동시에 전송하고자 할 때도 적용 가능하다. 적용 결과는 제안한 복수 트리경로에 의한 IP 멀티캐스팅 방법이 단-대-단 서비스 품질을 보장하는 데도 상당한 효과를 있음을 보인다. This paper proposes an IP(Internet Protocol) multicast routing method by multiple tree routes. Multiple trees, instead of a single tree, improve the quality of multicast services with nonlinear link cost and huge traffic demand. The proposed method adds tree routes until it satisfies target conditions, and it splits the multicast traffic demand into the chosen tree routes. We develop a mathematical model and optimal conditions for traffic splitting. The method works on the problems with many different simultaneous multicast traffic. Various experiments were carried and the results show that the new multicasting is fairly effective on end-to-end quality of services.

일반 교통망에서 브라이스 역설 발견 모형

박구현(Koohyun Park) 한국경영과학회 2007 한국경영과학회지 Vol.32 No.4

This study is for detecting the Braess Paradox by stable dynamics in general transportation networks. Stable dynamics, suggested by Nesterov and de Palma[18], is a new model which describes and provides a stable state of congestion in urban transportation networks. In comparison with user equilibrium model based on link latency function in analyzing transportation networks, stable dynamics requires few parameters and is coincident with intuitions and observations on the congestion. Therefore it is expected to be an useful analysis tool for transportation planners. The phenomenon that increasing capacity of a network, for example creating new links, may decrease its performance is called Braess Paradox. It has been studied intensively under user equilibrium model with link latency function since Braess[5] demonstrated a paradoxical example. However it is an open problem to detect the Braess Paradox under stable dynamics. In this study, we suggest a method to detect the Paradox in general networks under stable dynamics. In our model, we decide whether Braess Paradox will occur in a given network. We also find Braess links or Braess crosses if a network permits the paradox. We also show an example how to apply it in a network.

안정동력학 모형에서 최선 통행료 및 차선 통행료

박구현(Koohyun Park) 한국경영과학회 2009 한국경영과학회지 Vol.34 No.4

This study examined the first- and second-best pricing by stable dynamics in congested transportation networks. Stable dynamics, suggested by Nesterov and de Palma (2003), is a new model which describes and provides a stable state of congestion in urban transportation networks. The first-best pricing in user equilibrium models introduces user-equilibrium in the system-equilibrium by tolling the difference between the marginal social cost and the marginal private cost on each link. Nevertheless, the second-best pricing, which levies the toll on some, but not all, links, is relevant from the practical point of view. In comparison with the user equilibrium model, the stable dynamic model provides a solution equivalent to system-equilibrium if it is focused on link flows. Therefore the toll interval on each link, which keeps up the system-equilibrium, is more meaningful than the first-best pricing. In addition, the second-best pricing in stable dynamic models is the same as the first-best pricing since the toll interval is separately given by each link. As an effect of congestion pricing in stable dynamic models, we can remove the inefficiency of the network with inefficient Braess links by levying a toll on the Braess link. We present a numerical example applied to the network with 6 nodes and 9 links, including 2 Braess links.

안정동력학에 의한 가변수요 통행배정모형

박구현(Koohyun Park) 한국경영과학회 2009 韓國經營科學會誌 Vol.34 No.1

This study developed a variable demand traffic assignment model by stable dynamics. Stable dynamics, suggested by Nesterov and de Palma[19], is a new model which describes and provides a stable state of congestion in urban transportation networks. In comparison with the user equilibrium model, which is based on the arc travel time function in analyzing transportation networks, stable dynamics requires few parameters and is coincident with intuitions and observations on congestion. It is therefore expected to be a useful analysis tool for transportation planners. In this study, we generalize the stable dynamics into the model with variable demands. We suggest a three stage optimization model. In the first stage, we introduce critical travel times and dummy links and determine variable demands and link flows by applying an optimization problem to an extended network with the dummy links. Then we determine link travel times and path flows in the following stages. We present a numerical example of the application of the model to a given network.

지능망 서비스 제어 시스템의 위치 및 용량 결정에 관한 최적화 모델

박구현(Koohyun Park),신민수(Minsoo Shin),진병문(Byoung-Moon Chin) 한국통신학회 1990 한국통신학회 학술대회논문집 Vol.1990 No.8

평균/VaR 최적화 모형에 의한 전환사채 주식전환 비중 결정

박구현(Koohyun Park) 한국경영과학회 2013 經營 科學 Vol.30 No.3

In this study we suggested two optimization models to determine conversion weight of convertible bonds. The problem of this study is same as that of Park and Shim [1]. But this study used Value-at-Risk (VaR) for risk measurement instead of CVaR, Conditional-Value-at-Risk. In comparison with conventional Markowitz portfolio models, which use the variance of return, our models used VaR. In 1996, Basel Committee on Banking Supervision recommended VaR for portfolio risk measurement. But there are difficulties in solving optimization models including VaR. Benati and Rizzi [5] proved NP-hardness of general portfolio optimization problems including VaR. We adopted their approach. But we developed efficient algorithms with time complexity O(nlogn) or less for our models. We applied examples of our models to the convertible bond issued by a semiconductor company Hynix.

전환사채 주식전환을 위한 조건부 VaR 최적화

박구현(Koohyun Park),심은택(Eun-Tak Shim) 한국경영과학회 2011 經營 科學 Vol.28 No.2

In this study we suggested two optimization models to answer a question from an investor standpoint:how many convertible bonds should one convert, and how many keep? One model minimizes certain risk to the minimum required expected return, the other maximizes the expected return subject to the maximum acceptable risk. In comparison with Markowitz portfolio models, which use the variance of return, our models used Conditional Value-at-Risk(CVaR) for risk measurement. As a coherent measurement, CVaR overcomes the shortcomings of Value-at-Risk(VaR). But there are still difficulties in solving CVaR including optimization models. For this reason, we adopted Rockafellar and Uryasev’s[18, 19] approach. Then we could approximate the models as linear programming problems with scenarios. We also suggested to extend the models with credit risk, and applied examples of our models to Hynix 207CB, a convertible bond issued by the global semiconductor company Hynix.

WDM 수동 성형망에서의 클라이언트/서버 트래픽을 위한 TDM 스케쥴링 알고리즘

신홍섭(Hongseob Shin),박구현(Koohyun Park),노장래(Jang-Rai Roh) 한국정보과학회 1999 정보과학회논문지 : 시스템 및 이론 Vol.26 No.3

본 논문은 WDM(wavelength division multiplexing) 수동 성형망(passive star network)에서의 클라이언트/서버 트래픽을 위한 효율적인 TDM(time division multiplexing) 스케쥴링을 제시한다. 클라이언트/서버 트래픽은 서버 노드의 송, 수신 트래픽 양이 클라이언트 노드간의 트래픽 양에 비해 상대적으로 크다. 이러한 트래픽 특성 때문에, TT-FR 시스템의 TDM 스케쥴링 방법을 적용할 때에 TDM 프레임 길이가 길어지게 된다. 본 논문에서 제시하는 TDM 스케쥴링은 TT-TR 시스템 알고리즘이다. 제시하는 스케쥴링에서는 서버 노드의 송신 및 수신에 각각 하나씩의 파장을 먼저 할당한다. 따라서 각 클라이언트 노드는 송, 수신 모두를 위해 파장 조정이 요구된다. 이러한 새로운 TDM 스케쥴링은 클라이언트/서버 트래픽에 적합하고 기존의 방법보다 짧은 프레임 길이를 제공하며, 또한 TDM 프레임 길이의 하한을 새로이 제공한다. 제시하는 TT-TR 시스템 기반의 TDM 스케쥴링 알고리즘의 수행도 평가를 위해 다양한 시뮬레이션 분석을 수행하였다. This paper introduces an efficient TDM(tlme division multiplexing) scheduling algorithm for client/server traffic in WDM(wavelength division multiplexing) passive star networks. The amount of traffic transmitted or received by server is larger than the clients' traffic in a client/server traffic pattern. Because of this characteristic, the frame length is increased when the TT-FR system method of TDM scheduling is applied to the client/server traffic. The scheduling introduced in this paper is TT-TR system algorithm In this scheduling, two wavelengths are assigned to the server transmitting traffic and server receiving traffic, respectively. Consequently, each client have to tune its transmitter and receiver to transmit and receive data. This new TDM scheduling algorithm is suitable for the client/server traffic, and its frame length is shorter than that of previous method. This algorithm also provides a new lower bound of TDM frame length. For the performance analysis of this algorithm, various simulations are performed.

계층구조를 갖는 WDM 광전달망에서 파장할당 알고리즘

신용식(Yongsik Shin),이승민(Seungmin Lee),박구현(Koohyun Park) 한국통신학회 1997 한국통신학회 학술대회논문집 Vol.1997 No.11

트래픽 증가망에서의 최단경로설정: 송신 트래픽이 주어질 때

朴丘鉉 弘益大學校 科學技術硏究所 2004 科學技術硏究論文集 Vol.15 No.-

This paper introduces a shortest path routing algorithm in traffic-added networks where a linear traffic increase is introduced. This problem arises from routing in inter-networks with many different networks. The proposed method finds a minimum cost path for a given amount of traffic at the starting node. Since the amount of traffic at each arc is increased, the proposed method is different from the shortest path algorithms such as Dijkstra method. In this study, a mathematical model and optimality conditions were developed. A numerical example of the algorithm was reported.