• Proceedings of the Korea Information Processing Society Conference
• /
• 2005.05a
• /
• pp.1347-1350
• /
• 2005

본 논문에서는 에드혹 네트워크에서 트래픽이 분산되고 송신자 수의 증가에 강인한 멀티캐스팅 프로토콜을 제안한다. 본 논문에서 제안하는 지역 기반 트리를 이용한 멀티캐스팅 (RTMA)은 위치 정보를 이용하여 지역을 나누고 각 지역 안에서 멀티캐스트 그룹 멤버 간 공유 트리를 구성하고 지역 대표 노드를 선출하게 된다. 이후 멀티캐스트 그룹의 송신자를 포함하는 지역 공유 트리의 대표 노드들은 전체 지역 대표 노드를 대상으로 하는 송신자 기반 트리를 구성하여 지역 간 최단 트리를 만들게 된다. 이를 통하여 트래픽의 집중 현상을 방지하고 전체 송신자의 수가 아닌 송신자를 포함한 지역의 수 만큼만 트리를 생성하도록 하여 트리 유지를 위한 자원의 낭비를 줄일 수 있다.

• Journal of KIISE:Information Networking
• /
• v.30 no.3
• /
• pp.424-436
• /
• 2003

In this paper, we propose the Centroid-based Backbone Core Tree(CBCT) generation algorithm for the shared tree-based IP multicasting. The proposed algorithm is based on the Core Based Tree(CBT) protocol. Despite the advantages over the source-based trees in terms of scalability, the CBT protocol still has the following limitations; first, the optimal core router selection is very difficult, and second, the multicast traffic is concentrated near a core router. The Backbone Core Tree(BCT) protocol, as an extension of the CBT protocol has been proposed to overcome these limitations of the CBT Instead of selecting a specific core router for each multicast group, the BCT protocol forms a backbone network of candidate core routers which cooperate with one another to make multicast trees. However, the BCT protocol has not mentioned the way of selecting candidate core routers and how to connect them. The proposed CBCT generation algorithm employs the concepts of the minimum spanning tree and the centroid. For the performance evaluation of the proposed algorithm, we showed the performance comparison results for both of the CBT and CBCT protocols.

• Proceedings of the Korea Society for Simulation Conference
• /
• 2002.05a
• /
• pp.157-162
• /
• 2002

심화되는 경쟁환경에서 우위를 점하기 위해서는 제조 시스템의 수행도를 관리하고 개선하는 활동이 필요하다. 본 연구는 페트리넷(Petri net)과 시뮬레이션을 사용하여 제조 시스템 수행도의 하나인 산출율을 향상시키는 방법론을 제안하였다. 시뮬레이션은 제조 시스템의 수행도 평가에 많이 사용되나, 시스템의 개선방향에 대한 분석 능력이 취약하다. 한편, 페트리넷은 정확한 분석을 수행할 수 있으나, 대상 시스템이 복잡해짐에 따라 분석 능력이 현저히 감소된다. 본 연구에서는 페트리넷의 복잡도를 증가시키는 주요 원인 중 하나인 자원 공유를 분리된 페트리넷으로 표현함으로써 페트리넷의 분석 능력을 유지하고자 하였다. 또, 이를 바탕으로 자원 공유에 관련된 파라미터에 대한 회귀분석모형을 도출하였다. 이 회귀분석모형에 대한 최적화 과정에 의해 자원의 적절한 배치 혹은 제품비율에 대한 개선 방향을 얻어낼 수 있다. 이러한 페트리넷 기반 분석 방법과 시뮬레이션 방법을 동시에 사용함으로써 시뮬레이션 모형에서 산출율을 개선할 수 있으며, 실제 제조 시스템에 적용할 수 있을 것이다.

• The Journal of the Acoustical Society of Korea
• /
• v.18 no.6
• /
• pp.49-56
• /
• 1999

In many continuous speech recognition systems based on HMMs, decision tree-based state tying has been used for not only improving the robustness and accuracy of context dependent acoustic modeling but also synthesizing unseen models. To construct the phonetic decision tree, standard method performs one-level pruning using just single Gaussian triphone models. In this paper, two novel approaches, two-level decision tree and multi-mixture decision tree, are proposed to get better performance through more accurate acoustic modeling. Two-level decision tree performs two level pruning for the state tying and the mixture weight tying. Using the second level, the tied states can have different mixture weights based on the similarities in their phonetic contexts. In the second approach, phonetic decision tree continues to be updated with training sequence, mixture splitting and re-estimation. Multi-mixture Gaussian as well as single Gaussian models are used to construct the multi-mixture decision tree. Continuous speech recognition experiment using these approaches on BN-96 and WSJ5k data showed a reduction in word error rate comparing to the standard decision tree based system given similar number of tied states.

• The KIPS Transactions:PartC
• /
• v.10C no.7
• /
• pp.885-890
• /
• 2003

The Core Base Tree (CBT) multicast routing architecture is a multicast routing protocol for the internet. The CBT establishes a single shared tree for a multicast connection. The shared tree Is rooted at a center node called core. The location of the core may affect the cost and performance of the CBT. The core placement method requires the knowledge of the network topology In this Paper, we propose a simple and effective method for selecting the core. This method requires the distance vector information. in addition, we used results that calculated sample correlation coefficient. And then we select suitable routing algorithm according to member's arrangement states in muliticast group. we select core node that have minimum average cost or PIM-SM protocol is selected. The performance of this method is compared with several other methods by extensive simulations (i.e mean delay, maximum delay, and total cost). Our results shows that this method for Selecting Core is very effective.

• Journal of Digital Convergence
• /
• v.13 no.1
• /
• pp.243-248
• /
• 2015

Recognition model is not defined when you configure a model, Been added to the model after model building awareness, Model a model of the clustering due to lack of recognition models are generated by modeling is causes the degradation of the recognition rate. In order to improve decision tree state tying modeling using parameter estimation of Bayesian method. The parameter estimation method is proposed Bayesian method to navigate through the model from the results of the decision tree based on the tying state according to the maximum probability method to determine the recognition model. According to our experiments on the simulation data generated by adding noise to clean speech, the proposed clustering method error rate reduction of 1.29% compared with baseline model, which is slightly better performance than the existing approach.

• The KIPS Transactions:PartC
• /
• v.8C no.1
• /
• pp.60-67
• /
• 2001

Multicast Protocol for multimedia service on the Internet can be classified into two types, e.g., source based tree and shared tree according to difference of tree construction method. Shared tree based multicast is known to show outstanding results in the aspect of scalability than source based tree. Generally, There have been lots of researches on the method to satisfy QoS constraints through proper Rendezvous Point (RP) in the shared tree. In addition, as the multicast group members join and leave dynamically in the service time, RP of the shared tree should b be reselected for guranteeing Qos to new member, But, RP reselection method has not been considered generally as the solution to satisfy QoS C constraints. In this paper, new initial RP selection and RP reselection method are proposed, which utilize RTCP (Real Time Control Protocol) report packet fields. Proposed initial RP selection and RP reselection method use RTCP protocol which underlying multimedia application service So, the proposed method does not need any special process for collecting network information to calculate RP. New initial RP selection method s shows better performance than random and topology based one by 40-50% in simulation. Also, RP reselection method improves delay p performance by 50% after initial RP selection according to the member’s dynamicity.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.11 no.5
• /
• pp.101-106
• /
• 2011

In this paper, our goal is to design the controller which operates a manufacturing system, discrete event system, guaranteeing user specification. In specially, based on supervisory control theory W. M. Wonham, our work consists in performing the controller using Petri nets possible parallelism. In addition, we consider the problem of allocation for resources sharing of the issues to consider when designing using Petri net. The controller can be generated by synthesis of user specification model and plant model after giving the management for the resource sharing. This created controller can control the deadlock and starvation which can occur in the case with resource sharing. Previous studies proposed the Constrained Synchronous Reachability graph to synthesis of the two Petri nets models. Therefore, we provide the controlled system which forbids a deadlock and starvation using the result of previous studies.

• Journal of Internet Computing and Services
• /
• v.17 no.4
• /
• pp.69-77
• /
• 2016

Recently, the importance of visual analytics has been recognized in the field of business intelligence. From the view of business intelligence, visual analytics aims for acquiring valuable insights for decision making by interactively visualizing a variety of business information. In this paper, we propose a treemap-based method for visualizing work-sharing relationships among business process performers. A work-sharing relationship is established between two performers who jointly participate in a specific activity of a business process and is an important factor for understanding organizational structures and behaviors in a process-centric organization. To this end, we design and implement a treemap-based visualization tool for representing work-sharing relationships as well as basic hierarchical information in business processes. Finally, we evaluate usefulness of the proposed visualization tool through an operational example using XPDL (XML Process Definition Language) process models.

• The KIPS Transactions:PartD
• /
• v.8D no.5
• /
• pp.593-604
• /
• 2001

A database sharing system (DSS) refers to a system for high performance transaction processing. In the DSS, the processing nodes are coupled via a high speed network and share a common database at the disk level. Each node has a local memory and a separate copy of operating system. To reduce the number of disk accesses, the node caches data pages and index pages in its memory buffer. In general, B-tree index pages are accessed more often and thus cached at more processing nodes, than their corresponding data pages. There are also complicated operations in the B-tree such as Fetch, Fetch Next, Insertion and Deletion. Therefore, an efficient cache consistency scheme supporting high level concurrency is required. In this paper, we propose cache consistency schemes using identifiers of index pages and page_LSN of leaf page. The propose schemes can improve the system throughput by reducing the required message traffic between nodes and index re-traversal.