• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제8권3호
• /
• pp.762-777
• /
• 2014

With the increasing usage of cloud applications such as MapReduce and social networking, the amount of data traffic in data center networks continues to grow. Moreover, these appli-cations follow the incast traffic pattern, where a large burst of traffic sent by a number of senders, accumulates simultaneously at the shallow-buffered data center switches. This causes severe packet losses. The currently deployed TCP is custom-tailored for the wide-area Internet. This causes cloud applications to suffer long completion times towing to the packet losses, and hence, results in a poor quality of service. An Explicit Congestion Notification (ECN)-based approach is an attractive solution that conservatively adjusts to the network congestion in advance. This legacy approach, however, lacks scalability in terms of the number of flows. In this paper, we reveal the primary cause of the scalability issue through analysis, and propose a new congestion-control algorithm called FaST. FaST employs a novel, virtual congestion window to conduct fine-grained congestion control that results in improved scalability. Fur-thermore, FaST is easy to deploy since it requires only a few software modifications at the server-side. Through ns-3 simulations, we show that FaST improves the scalability of data center networks compared with the existing approaches.

• 전기전자학회논문지
• /
• 제16권1호
• /
• pp.45-50
• /
• 2012

본 논문은 설계 흐름의 초기 단계에서 SoC의 라우팅 밀집도를 빠르고 정확하게 예측하는 방법론을 제안한다. 라우팅 과정에서 발생하는 과도한 밀집 현상은 라우팅 실패를 야기하고, 물리 설계를 처음부터 다시하게 되는 불필요한 시간을 소모하게 한다. 설계 초기단계에서 라우팅 밀집도를 정확하게 예측하는 것은 성공적인 물리 설계를 이끌어 내며, 전체 설계 시간에 소모되는 비용을 최소화시킨다. 제안된 방법은 블록 수준 플로어플랜 단계에서 블록 간/블록 내부 인터커넥트, 가상으로 합성된 파워/클록 네트워크를 사용해서 정교한 라우팅 밀집도를 예측한다.

• 전기학회논문지P
• /
• 제63권3호
• /
• pp.131-137
• /
• 2014

It is expected that there will be delay of scheduled transmission network reinforcement and huge investment of renewable energy resources in Korea. As transmission capacity expansion delayed, supplying power to Seoul metropolitan area will not be increased as scheduled. In addition, uncertain renewable energy out of Seoul metropolitan area can cause transmission congestion in the future power system. These two combining effects will make the difference in locational marginal prices(LMP) and congestion costs increase. In that sense, this paper will analyze how much the congestion costs for Korea power system are incurred in the future power system. Most of previous approaches to analyze the congestion costs for electric power system are based on the optimal power flow model which cannot deal with hourly variation of power system. However, this study attempted to perform the analysis using market simulation model(M-Core) which has the capability of analyzing the hourly power generation cost and power transmission capacity, and market prices by region. As a result, we can estimate the congestion costs of future power system considering the uncertainty of renewable energy and transmission capacity.

• 산업공학
• /
• 제23권4호
• /
• pp.349-356
• /
• 2010

The role of Intelligent Transportation System (ITS) is to efficiently manipulate the traffic flow and reduce the cost in logistics by using the state of the art technologies which combine telecommunication, sensor, and control technology. Especially, the hardware part of ITS is rapidly adapting to the up-to-date techniques in GPS and telematics to provide essential raw data to the controllers. However, the software part of ITS needs more sophisticated techniques to take care of vast amount of on-line data to be analyzed by the controller for their decision makings. In this paper, the authors develop a traffic congestion prediction model based on several different parameters from the sensory data captured in the Vehicle Detection System (VDS). This model uses the neural network technology in analyzing the traffic flow and predicting the traffic congestion in the designated area. This model also validates the results by analyzing the errors between actual traffic data and prediction program.

• 정보처리학회논문지C
• /
• 제13C권3호
• /
• pp.353-358
• /
• 2006

TCP 구현의 하나인 Vegas는 패킷의 유실을 망의 혼잡으로 인지하는 Reno와 달리 RTT(Round Trip Time) 측정값을 바탕으로 혼잡을 인지하며, 이를 이용하여 윈도우 크기 등 혼잡 제어를 위한 주요 인자를 결정한다. 그러나, Vegas의 혼잡 회피 방안이 TCP 패킷 경로의 비대칭적 특성을 제대로 반영하지 못하며, 이것은 양 방향(순방향, 역방향) 패킷 전송 상태를 반영하는 RTT 측정값을 순방향 경로의 상태 해석에 이용하기 때문이다. RTT는 패킷의 왕복 시간만을 측정하기 때문에 패킷의 송수신시 순방향과 역방향에서 어느 정도의 혼잡이 발생하였는지 알 수 없다. 본 논문에서는 리눅스 커널 2.6의 TCP 소스에서 RTT 측정값으로 혼잡도를 측정하는 기존의 Vegas 혼잡 제어 알고리즘을 수정하여 순방향 경로의 혼잡과 역방향 경로의 혼잡을 구별할 수 있는 새로운 Vegas 혼잡 제어 알고리즘을 설계하고 구현하여 그 성능을 분석하였다.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2008년도 학술대회 논문집 정보 및 제어부문
• /
• pp.544-545
• /
• 2008

무선 통신 기술의 발달로 최근 무선 device 들로 이루어진 ad-hoc network이 주요 issue가 되고 있으며, ad-hoc network에서의 성능이 중요한 관심사가 되고 있다. 그러나 유선망에 최적화되도록 설계된 TCP 는 무선망이 가지는 링크의 불안정함으로 인한 손실을 네트워크 혼잡에 의한 손실로 인식하여 혼잡 제어 메커니즘이 수행함으로써 전송률을 줄이게 되므로 전체적으로 네크웍의 성능이 저하되는 문제점을 가진다. 본 논문에서는 무선망에서 TCP 성능을 향상시키기 위한 새로운 congestion algorithm을 제안하고 그 성능을 평가하였다. 성능 확인 결과 기존 TCP-vegas에 비해 더 나은 성능을 보임을 알 수 있다.

• 한국통신학회논문지
• /
• 제22권2호
• /
• pp.259-267
• /
• 1997

The advent of B-ISDN using ATM(asynchronous transfer mode) made possible a variety of new multimedia services, however it also created a problem of congestion control due to bursty nature of various traffic sources. To tackle this problem, UPC/NPC(user parameter control/network parameter control) have been actively studied and DRLB(dynamic rate leaky bucket) algorithm, in which the token generation rate is changed according to states of data source andbuffer occupancy, is a good example of the UPC/NPC. However, the DRLB algorithm has drawbacks of low efficiency and difficult real-time implementation for bursty traffic sources because the determination of token generation rate in the algorithm is based on the present state of network. In this paper, we propose a more plastic and effective congestion control algorithm by combining the DRLB algorithm and neural network based prediction to remedy the drawbacks of the DRLB algorithm, and verify the efficacy of the proposed method by computer simulations.

• 디지털산업정보학회논문지
• /
• 제8권2호
• /
• pp.81-94
• /
• 2012

This paper suggests PARED algorithm, a modified RED algorithm, that actively reacts to dynamic changes in network to apply packet drop probability flexibly. The main idea of PARED algorithm is that it compares the target queue length to the average queue length which is the criterion of changes in packet drop probability and feeds the gap into packet drop probability. That is, when the difference between the average queue length and the target queue length is great, it reflects as much as the difference in packet drop probability, and reflects little when the difference is little. By doing so, packet drop probability could be actively controled and effectively dealt with in the network traffic situation. To evaluate the performance of the suggested algorithm, we conducted simulations by changing network traffic into a dynamic stat. At the experiments, the suggested algorithm was compared to the existing RED one and then to ARED one that provided the basic idea for this algorithm. The results proved that the suggested PARED algorithm is superior to the existing algorithms.

• ETRI Journal
• /
• 제36권6호
• /
• pp.960-967
• /
• 2014

In this paper, an analytical framework is proposed for the optimization of network performance through joint congestion control, channel allocation, rate allocation, power control, scheduling, and routing with the consideration of fairness in multi-channel wireless multihop networks. More specifically, the framework models the network by a generalized network utility maximization (NUM) problem under an elastic link data rate and power constraints. Using the dual decomposition technique, the NUM problem is decomposed into four subproblems - flow control; next-hop routing; rate allocation and scheduling; power control; and channel allocation - and finally solved by a low-complexity distributed method. Simulation results show that the proposed distributed algorithm significantly improves the network throughput and energy efficiency compared with previous algorithms.

• 한국통신학회논문지
• /
• 제31권8B호
• /
• pp.764-768
• /
• 2006

본 논문에서는 대용량의 대역폭을 사용하는 다양한 멀티미디어 서비스에서 발생할 수 있는 네트워크 혼잡 문제를 효율적으로 제어하며, 동시에 우선순위가 높은 서비스의 QoS를 보장하는 적응적 온라인 가격결정 방법을 제안하였다. 이 방법은 현재 네트워크의 트래픽 상황을 기반으로 하여 적응성과 유연성을 제공하는 실시간 온라인기법을 기반으로 수행되기 때문에 실제 네트워크 운영에 적용하기가 용이하다. 컴퓨터 시뮬레이션을 통하여 본 논문이 제안한 방법이 다양한 네트워크 트래픽 상황에서 우수한 성능을 가지는 것을 확인할 수 있었다.