• Proceedings of the KIEE Conference
• /
• 2007.10a
• /
• pp.81-82
• /
• 2007

자율이동로봇을 위한 새로운 경로 계획 방법을 제안한다. 실시간으로 장애물을 피하고, 목표 지점까지의 최단 경로를 생성하여 유용성을 극대화시키기 위하여 방법을 다룬다. 본 논문에서는 효율적인 경로 계획방법으로 초기 경로를 생성하고, 생성된 경로를 개선하는 방법을 제안한다. 초기 경로는 그래프 기반 방법인 골격선 그래프와 탐색방법으로 딕스트라(Dijkstra) 알고리즘을 사용한다. 초기 경로에 대해 동적 프로그래밍 알고리즘을 이용하여 최단거리에 가깝게 경로를 개선한다. 시뮬레이션을 통해 제안하는 방법의 성능을 검증한다.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 1999.10a
• /
• pp.97-97
• /
• 1999

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.10 no.5
• /
• pp.265-270
• /
• 2010

Reinforcement learning performs learning through interacting with trial-and-error in dynamic environment. Therefore, in dynamic environment, reinforcement learning method like TD-learning and TD(${\lambda}$)-learning are faster in learning than the conventional stochastic learning method. However, because many of the proposed reinforcement learning algorithms are given the reinforcement value only when the learning agent has reached its goal state, most of the reinforcement algorithms converge to the optimal solution too slowly. In this paper, we present GDRLS algorithm for finding the shortest path faster in a maze environment. GDRLS is select the candidate states that can guide the shortest path in maze environment, and learn only the candidate states to find the shortest path. Through experiments, we can see that GDRLS can search the shortest path faster than TD-learning and TD(${\lambda}$)-learning in maze environment.

• Proceedings of the Korean Society of Marine Engineers Conference
• /
• 2005.11a
• /
• pp.206-207
• /
• 2005

해상에서 인터넷을 사용하는 데 많은 기술적, 경제적 제약이 따른다. 육상의 인터넷 서비스를 바다 위에서 실현하기 위해서는 기지국과 같은 중앙 통제 시스템이 없는 MANET을 기반으로 하는 해상통신망이 새롭게 구축되어야 한다. 본 논문에서는 해상환경과 선박 내 단말장치의 특성을 최대한 고려한 해상 MANET 모델을 새롭게 설계하였다. 원양에서 선박들이 주로 주어진 항로를 기준으로 항해하는 특성을 고려하여, 개별적 이동노드(선박)를 대상으로 경로를 찾는 것이 아니라 정적 정보인 항로와 동적 정보인 선박을 복합적으로 활용하여 최단경로를 찾아내는 복합적 항로기반 라우팅(HCBR) 프로토콜을 제안하였다. 지리 정보를 이용하는 LAR 프로토콜과 성능 비교에서 경로 발견 시 제어패킷을 전혀 쓰지 않으면서 최단 경로를 훨씬 더 잘 찾아내는 특성을 실험적으로 파악하였다.

• The KIPS Transactions:PartA
• /
• v.12A no.6 s.96
• /
• pp.557-562
• /
• 2005

Throughput and latency of interconnection network are important factors of the performance of multiprocessor systems. The dual-link CC-NUMA architecture using point-to-point unidirectional link is one of the popular structures in high-end commercial systems. In terms of optimal path between nodes, several paths exist with the optimal hop count by its native multi-path structure. Furthermore, transaction latency between nodes is affected by congestion of links on the transaction path. Hence the transaction latency may get worse if the transactions make a hot spot on some links. In this paper, I propose a dynamic transaction routing algorithm that maintains the balanced link utilization with the optimal path length, and I compare the performance with the fixed path method on the dual-link CC-NUMA systems. By the proposed method, the link competition is alleviated by the real-time path selection, and consequently, dynamic transaction algorithm shows a better performance. The program-driven simulation results show $1{\~}10\%$ improved fluctuation of link utilization, $1{\~}3\%$ enhanced acquirement of link, and $1{\~}6\%$ improved system performance.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2012.05a
• /
• pp.234-235
• /
• 2012

본 논문에서는 개미 집단 최적화(Ant Colony Optimization, ACO) 알고리즘을 적용한 감속률에 따른 동적 상황인식 경로 선정 방법을 제안한다. 최근 ITS(Intelligent Transportation Systems)의 개발과 함께 차량용 내비게이션의 실시간 교통 정보를 이용하는 수요가 급증하면서, 경로탐색의 중요성이 더욱 가속화되고 있다. 현재 차량용 내비게이션은 멀티미디어 및 정보통신 기술의 결합과 함께 다양한 기능 및 정보를 사용자에게 제공하고 있으며, 이러한 경로탐색 알고리즘은 교통시스템, 통신 네트워크, 운송 시스템 등 다양한 분야에 적용되고 있다. 본 논문에서는 감속률에 따른 동적 상황인식 경로 선정 방법을 제안함으로써, 최단 시간 및 최소 비용의 정보를 제공해 줄 뿐만 아니라 교통정체로 인한 사회적 비용 감소의 효과를 가져다 줄 것으로 기대한다.

• Journal of Korean Society of Transportation
• /
• v.22 no.7 s.78
• /
• pp.91-98
• /
• 2004

This study focuses on the development of dynamic assignment for evaluation and application for dynamic route guidance strategy. Travelers are classified according to information contents which they received pre/on trip. The first group have no traffic information, so they travel with fixed route. The second group have real-time shortest path and travel according to it. The last group have car navigation system or individual method(cellular phone, PDA-two way communication available) for traffic information on trip. And then they are assigned in accordance with the proposed multiclass dynamic assignment model. At this time the last group is gotten under control with DFS(decentralized feedback strategy). In use of this Process we expect that various traffic information strategy can be tested and also be the key factor for success of ITS, location of VMS(variable message sign), cycle of information, area of traffic information, etc).

• Journal of Digital Convergence
• /
• v.10 no.4
• /
• pp.223-228
• /
• 2012

There are many algorithms to improve the network traffic and to avoid losing packets in the network. This paper proposed determining the shortest paths for hops, which are in the middle of the source and destination. The shortest path in this paper means the fastest path between them. Recently, dynamic routing algorithm is currently used now but this paper suggests the fastest paths between the source and the destination is by using the record of the network traffic history. People are using the networking and the network traffic is always corresponding to how many people use the networking in specific time. Therefore, I can predict the network condition by referring to the history of network traffic record, and then the shortest path can be produced without using RIP too much. It will be helpful to improve the network traffic.

• The KIPS Transactions:PartB
• /
• v.9B no.2
• /
• pp.155-162
• /
• 2002

Reinforcement learning is a learning method that uses trial-and-error to perform Learning by interacting with dynamic environments. It is classified into online reinforcement learning and delayed reinforcement learning. In this paper, we propose an online reinforcement learning system (ONRELS : Outline REinforcement Learning System). ONRELS updates the estimate-value about all the selectable (state, action) pairs before making state-transition at the current state. The ONRELS learns by interacting with the compressed environments through trial-and-error after it compresses the state space of the mage environments. Through experiments, we can see that ONRELS can search the shortest path faster than Q-learning using TD-ewor and $Q(\lambda{)}$-learning using $TD(\lambda{)}$ in the maze environments.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.10 no.7
• /
• pp.1312-1318
• /
• 2006

One of the common services of telematics is the car navigation that finds the shortest path from source to target. Until now, some routing algorithms of the car navigation do not consider the real-time traffic information and use the static shortest path algorithm. In this paper, we prosed the method to predict the traffic flow in the future. This prediction combines two methods. The former is an accumulated speed pattern, which means the analysis results for all past speeds of each road by classfying the same day and the same time inteval. The latter is the Kalman filter. We predicted the traffic flows of each segment by combining the two methods. By experiment, we showed our algorithm gave better precise predicition than only using accumulated speed pattern that is used commonly. The result can be applied to the car navigation to support a dynamic shortest path. In addition, it can give users the travel information to avoid the traffic congestion areas.