• Proceedings of the Korean Society of Computer Information Conference
• /
• 2012.01a
• /
• pp.5-6
• /
• 2012

본 논문에서는 가상머신에서 수행되는 어플리케이션의 서비스 품질을 보장하고, 클라우드 클러스터의 운영비용을 절감시킬 수 있는 자원 할당 감축기법을 제안한다. 이 기법은 가상머신의 자원 사용량 변화 추세를 분석하고 이를 토대로 확률적인 접근을 사용하여 새로운 자원 할당 감축량을 결정한다. 가상머신의 자원 사용량이 할당량을 초과하면 가상머신의 이주가 필요하다. 이때 발생하는 다운타임동안 가상머신의 어플리케이션은 서비스를 수행할 수 없기 때문에 클라우드 시스템의 성능이 저하된다. 따라서 성능평가에서는 가상머신의 자원 사용량이 할당량을 초과하는 횟수를 측정하여 Greedy 기법과 비교 평가함으로써 제안한 기법이 자원 할당 감축에 우수함을 검증한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2010.04a
• /
• pp.733-736
• /
• 2010

VANET(Vehicular Ad-hoc Network)은 노드를 차량으로 가정하는 개념의 MANET(Mobile Ad-hoc Network)로서 노드의 빠른 이동으로 인해 급격한 토폴로지의 변화가 일어난다. 하지만 차량 노드의 이동은 도로 상에서 제한되어 있기 때문에 토폴로지에 대한 상대적인 예측 가능성을 가지고 있다. 이는 교통이 혼잡한 도로 환경에서 그리디 기법을 이용하여 다음 홉을 결정할 때 보다 높은 정확성을 제공할 수 있어 경유 노드의 수와 포워딩 실패를 최소화한다. 본 논문은 위기 정보와 운전 시스템 정보를 기반으로 하는 차량 간 통신 라우팅 최적화 기법을 제안하고 기존의 GPSR(Greedy Perimeter Stateless Routing) 기법과 SAR(Spatial Aware Routing) 기법과의 비교를 통해 효율성과 신뢰성의 향상을 증명하였다.

• KEPCO Journal on Electric Power and Energy
• /
• v.2 no.3
• /
• pp.391-395
• /
• 2016

In the last few decades, global electricity consumption has dramatically increased and has become drastically fluctuating and uncertain causing blackout. Due to the unexpected peak electricity demand, we need significant electricity supply. The solutions to these problems are smart grid system which is envisioned as future power system. Smart grid system can reduce electricity peak demand and induce effective electricity consumption through various price policies, demand response (DR) control methodologies, and state-of-the-art smart equipments in order to optimize electricity resource usage in an intelligent fashion. Demand response (DR) is one of the key technologies to enable smart grid. In this paper, we propose greedy technique for demand response smart grid system. The proposed scheme focuses on minimizing electricity bills, preventing system blackout and sacrificing user convenience.

• Journal of KIISE:Computer Systems and Theory
• /
• v.36 no.5
• /
• pp.422-428
• /
• 2009

NAND flash memory is widely used in embedded systems because of many attractive features, such as small size, light weight, low power consumption and fast access speed. However, it requires garbage collection, which includes erase operations. Erase operation is slower than other operations. Further, a block has a limited erase lifetime (typically 100,000) after which a block becomes unusable. The proposed garbage collection policy focuses on minimizing the total number of erase operations, the deviation value of each block and the garbage collection time. NAND flash memory consists of pages of three types, such as valid pages, invalid pages and free pages. In order to achieve above goals, we use a page ratio to decide when to do garbage collection and to select the target victimblock. Additionally, we implement allocating method and group management method. Simulation results show that the proposed policy performs better than Greedy or CAT with the maximum rate 85% of reduction in the deviation value of the erase operations and 6% reduction in garbage collection time.

• Journal of Korea Society of Industrial Information Systems
• /
• v.15 no.4
• /
• pp.7-15
• /
• 2010

This study proposes a rad o resource management scheme to maximize the performance of the LTE(Long Term Evolution) uplink, using SC-FDMA(Single Carrier-Frequency Division Multiple Access). Rather than the single-cell SC-FDMA system the existing studies are mainly concerning, this study focuses on multi-cell system which needs considering the interaction among cells. Radio resource management is divided into two phases, planning and operation phases. The former is for the master eNB(e-NodeB) to allocate RBs(radio bearer) to eNB, the latter for eNB to assign RBs to the mobiles in the cell. For each phase, an optimization model and greedy algorithm are proposed. Optimization models aim to maximize the system performance while satisfying the constraints for both QoS and RB continuity. The greedy algorithms, like generic ones, move from a solution to a neighboring one having the best objective value among neighboring ones. From the numerous numerical experiments, the performance and characteristics of the algorithms are analyzed. This study is expected to play a volunteering role in radio resource management for the multi-cell SC-FDMA system.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2012.11a
• /
• pp.114-116
• /
• 2012

Sparse한 신호 복원 방법으로 underdetermined system에서 greed 알고리듬은 간결함과 낮은 복잡도로 인해 활발히 연구되고 있다. 이에 본 논문은 기존 greed 알고리듬 기법에서 iteration 마다 다중 경로를 이용하여 스파스 신호를 복원하는 개선된 알고리듬을 제안한다. 모의 실험을 통해 제안된 알고리듬이 기존의 greedy 알고리듬보다 좋은 복원 성능을 가짐을 확인할 수 있다.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.5 no.1
• /
• pp.33-44
• /
• 2012

A class of recurrent neural networks is proposed and proven to be capable of identifying any discrete-time dynamical system. The applications of the proposed network are addressed in the encoding, identification, and extraction of finite state automata. Simulation results show that the identification of finite state automata using the proposed network, trained by the hybrid greedy simulated annealing with a modified error function in the learning stage, exhibits generally better performance than other conventional identification schemes.

• The Journal of the Korea Contents Association
• /
• v.9 no.1
• /
• pp.98-105
• /
• 2009

In this paper, new data centric storage that is dynamically adapted to the change of work load is proposed. The proposed data centric storage distributes the load of hot spot area by using multilevel grid technique. Also, the proposed method is able to use existing routing protocol such as GPSR (Greedy Perimeter Stateless Routing) with small changes. Through simulation the proposed method enhances the lifetime of sensor networks over one of the state-of-the-art data centric storages. We implement the proposed method based on a operating system for sensor networks, and evaluate the performance through running based on a simulation tool.

• The Journal of the Korea Contents Association
• /
• v.10 no.5
• /
• pp.1-9
• /
• 2010

This paper propose a novel scheme called selective evaluation to improve convergence of ACO (ant colony optimization) for feature selection. The scheme cutdown the computational load by excluding the evaluation of unnecessary or less promising candidate solutions. The scheme is realizable in ACO due to the valuable information, pheromone trail which helps identify those solutions. With the aim of checking applicability of algorithms according to problem size, we analyze the timing requirements of three popular feature selection algorithms, greedy algorithm, genetic algorithm, and ant colony optimization. For a rigorous timing analysis, we adopt the concept of atomic operation. Experimental results showed that the ACO with selective evaluation was promising both in timing requirement and recognition performance.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.49 no.2
• /
• pp.122-129
• /
• 2012

As a greedy algorithm reconstructing the sparse signal from underdetermined system, orthogonal matching pursuit (OMP) algorithm has received much attention in recent years. In this paper, we present an extension of OMP for pursuing efficiency of the index selection. Our approach, referred to as generalized OMP (gOMP), is literally a generalization of the OMP in the sense that multiple (N) columns are identified per step. Using the restricted isometry property (RIP), we derive the condition for gOMP to recover the sparse signal exactly. The gOMP guarantees to reconstruct sparse signal when the sensing matrix satisfies the RIP constant ${\delta}_{NK}$ < $\frac{\sqrt{N}}{\sqrt{K}+2\sqrt{N}}$. In addition, we show recovery performance and the reduced number of iteration required to recover the sparse signal.